Electrocardiographic Findings in Patients Hospitalized With COVID-19: Retrospective Study
Abstract
Background: COVID-19 can involve the heart, which is associated with an increase in morbidity and mortality and can detected by electrocardiogram (ECG). So, in this study, we assess ECG changes in COVID-19-infected patients during hospitalization.
Methods: In this study, we examined the ECGs of 474 patients with COVID-19 positive at Birjand Vali-Asr Hospital (east of Iran). All the ECGs were taken by a single device and interpreted by a cardiologist. Demographic information, past medical history, and severity of COVID disease (in terms of oxygen saturation, respiratory distress, and need for emergency care) were entered in the checklist. Data were entered into SPSS Version 22 and analyzed. A p value of ≤ 0.05 was considered as a significant level.
Results: The patient’s mean age was 57.13 ± 18.7 years, and 55% of them were men; 428 (89.5%) patients survived and 49 (10.5%) died. About 78% of patients had abnormal ECG, which was higher in deceased patients significantly (91.8%, p = 0.03). Most frequent abnormalities include 40.3% of dysrhythmia (sinus tachycardia, atrial fibrillation, and sinus bradycardia), 30.1% of abnormal rate (tachycardia, bradycardia), and 30% of low-voltage ECG. Statistical analysis showed that after multivariable logistic regression, tachycardia (OR = 2.65 [1.2–5.8]; p = 0.015) and ST-segment elevation (OR = 2.9 [1.2–7.2]; p = 0.023) were directly related to the disease severity and mortality.
Conclusion: ECG changes are very common in COVID-19 patients, especially rate and rhythm changes. Tachycardia and ST segment elevation were associated with mortality. ECG changes are very common in COVID-19 patients, especially changes in rate and rhythm. Tachycardia and ST segment elevation were associated with mortality. Since the ECG is a simple, cheap, and safe diagnostic method, it is necessary to take an initial ECG from the patient in acute upper respiratory infections such as COVID-19, at least for patients with severe disease. In our study, the significant point was the presence of low voltage in ECG in one-third of the patients with COVID-19, which requires further research in this field.
1. Introduction
COVID-19 is causing a pandemic outbreak of respiratory disease and is spreading all over the world after December 2019 [1]. The organ that is most commonly affected in COVID-19 is the lung, which manifests as acute pneumonia. One of the most important organs involved in COVID-19 is the cardiovascular system. Cardiovascular involvement is very heterogeneous, including acute congestive heart failure, stress-induced cardiomyopathy, cardiogenic shock, acute myocarditis, even in fulminant form, acute coronary syndrome, thromboembolic events such as acute pulmonary embolism and arrhythmia [2]. Studies have shown that the mortality of COVID-19 patients with cardiovascular involvement is very high. Also, patients who had a previous history of cardiac disease or cardiac risk factors such as hypertension, diabetes mellitus, and obesity have a higher mortality rate than other people [3].
The cardiac involvement of patients is caused by several factors, including direct damage to the myocardium by entering the virus into the myocytes, direct penetration of SARS-CoV-2 into endothelium and endothelial dysfunction, activation and release of immune cells and proinflammatory cytokines, followed by a vigorous immune response, instability, and rupture of atherosclerotic plaque, creating a prethrombotic state through the release of thromboxane, von Willebrand factor, and plasminogen activator 1; adrenergic stimulation is caused by fever and hypoxemia and increased myocardial oxygen demand, and increase in the angiotensin II level following the entry of SARS-CoV-2 into cardiomyocytes through ACE2 [4]. The side effects of drugs used in the treatment of COVID-19, electrolyte disorders such as hypokalemia, hypoxia, and acidosis also play a role in the heart complications of COVID-19 [5].
One of the simple and cheap diagnostic methods of heart diseases is the electrocardiogram (ECG), which was used in the diagnosis of cardiac involvement of COVID-19 patients. The ECG can help in emergencies, determine the patient’s prognosis, and identify acute or chronic cardiac disease through various abnormal findings; it can be rapidly performed without inducing significant exposure to COVID-19. The ECG abnormalities in these patients included arrhythmia, P-R segment depression, ST elevation, atrio-ventricular block (AVB), QT prolongation, pseudo-infarct pattern, bradyarrhythmia, ventricular tachycardia, Brugada-like pattern, RBBB, and RV strain pattern [6–9]. Cardiac tachycardia includes sinus tachycardia, atrial fibrillation, and ventricular tachycardia. The mechanism of their creation is hypoxia, hyper-adrenergic state, autonomic dysfunction, direct viral endothelial damage, acid-base imbalance, metabolic and electrolyte abnormalities, myocarditis, and antiarrhythmic drugs [6–8]. Sinus bradycardia in COVID can be due to various causes, such as high levels of cytokines, direct toxicity of SARS-CoV-2 to the sinoatrial node (SAN) or atrioventricular node (AVN), parasympathetic overdrive, and the use of some drugs such as remdesivir or steroids [9]. Inflammatory effects of dangerous cytokines such as interleukin (IL)-1 and IL-6 and tumor necrosis factor alpha on AVN conduction system can be the cause of AVB block in COVID [10]. ST segment elevation or depression is commonly seen as a result of direct myocardial injury in the acute phase of the disease, myocarditis, acute myocardial infarction, and acute pericarditis. Acute respiratory failure also causes QRS axis deviation to the right (RAD), negative T in leads V1–V3, RBBB (RV strain pattern), and long QT interval [8]. QTc interval prolongation may be due to inflammation, renal dysfunction, and diarrhea, imbalance of intravascular fluids, and prescription of drugs (antivirals, antibiotics, and antiarrhythmics). Electrolyte disorders, especially hypokalemia, hypomagnesemia, and hypophosphatemia, can also prolong the QT interval, which is the cause of serious arrhythmias such as Torsade de Point [11].
ECG changes correlated with adverse clinical outcomes include acute respiratory distress syndrome, mechanical ventilation, intensive care unit (ICU) admission, renal failure, and in-hospital mortality [12–14]. In one study, the relationship among ST-T changes, sinus tachycardia, and atrial fibrillation with the severity of COVID disease has been seen, but after adjusting the effect of confounding variables, only atrial fibrillation and sinus tachycardia were independently related to the need for ventilation and mortality in COVID patients [12]. However, in another study, tachycardia, bundle branch block, ST elevation, fragmented QRS, QT prolongation, and premature atrial and ventricular complexes were related to mortality, and after multivariate analysis, the effects of these factors remained [13].
In another study, the relationship between ECG changes and mortality in COVID patients without heart disease, hospitalized in ICU, was investigated. The results showed that longer QTc, Tpe interval, and higher Tpe/QT and Tpe/QTc ratios are associated with higher mortality [14].
In this study, we investigated the ECG changes of hospitalized COVID patients. Our aim is to compare the ECG changes in different age groups, sex, disease severity, comorbidities, and in-hospital mortality.
2. Methods
2.1. Study Population
This is a retrospective study of infected COVID-19 patients, hospitalized from September 23, 2021, to January 19, 2022, in Birjand Vali-Asr Hospital of east of Iran. The criteria for entering the study were over 18 years of age, a positive PCR test for COVID-19, and having an ECG at the time of hospitalization.
2.2. Electrocardiographic Evaluation
ECGs were taken in a standard way (speed of 25 mm/s and voltage of 10 mm/mv in normal limb positioning). All ECGs are taken by a medical ECONET device, and the device is calibrated every 6 months. In order to avoid interpersonal errors, all ECGs were interpreted by a cardiologist in a standard way. The ECG definitions used in this study are tachycardia (heart rate ≥ 100 with any rhythm), bradycardia (heart rate < 60 with any rhythm), dysrhythmia (any reported rhythm except normal sinus rhythm), and repolarization abnormality (ST-segment elevation or ST-segment depression or T wave inversion), low voltage (QRS amplitude of < 5 mm in limb leads and/or < 10 mm in precordial leads), and abnormal ECG (including any ECG with pathologic point). Patients′ disease severity according to the Spo2 saturation in room air, respiratory distress, and need for emergency care were divided to severe, nonsevere, and dead.
2.3. Statistical Analysis
All conducted statistical tests were done by SPSS version 22.0. Quantitative variables were reported using mean and standard deviation, and qualitative variables were reported using frequency or percentage. If the variables had a normal distribution, the t-test was used; otherwise, the Mann–Whitney test was used. Chi-square test and Fisher’s exact test were also used to determine the relationship between qualitative variables. The association between ECG abnormalities and in-hospital mortality was assessed using univariate and multivariate logistic regression analyses. Odds ratio (OR) is expressed with a 95% confidence interval. p value ≤ 0.05 was considered significant.
3. Result
Our study was conducted on 474 COVID-19 positive patients from September 23, 2021, to January 19, 2022. The patient’s mean age was 57.13 ± 18.7 years old; 54.8% of them were men and 428 (89.5%) patients survived and 49 (10.5%) died.
As reported in Table 1, 78% of ECGs were abnormal, and the most frequent abnormalities include dysrhythmia (40.3%), rate abnormality (30.1%), and low-voltage ECG (30%). Increased heart rate, inverted T wave, and low-voltage ECG, were significantly higher in women and abnormal P wave and axis deviation were higher in men (Table 1).
| Variable | Female (n = 214) | Male (n = 260) | p value | Overall | |
|---|---|---|---|---|---|
| Rate (mean)/min | 91.6 ± 18.1 | 87.5 ± 20.1 | 0.022∗ | 89.4 ± 19.34 | |
| Rate | NL | 155 (72.4%) | 177 (67.8%) | 0.275 | 332 (69.9%) |
| Tachycardia | 54 (25.2%) | 65 (24.9%) | 119 (25%) | ||
| Bradycardia | 5 (2.3%) | 19 (7.3%) | 24 (5.1%) | ||
| Rhythm | NL | 127 (59.6%) | 154 (59.7%) | 0.988 | 281 (59.7%) |
| Sinus tachycardia | 68 (31.9%) | 74 (28.7%) | 142 (30.1%) | ||
| Sinus bradycardia | 9 (4.2%) | 20 (7.8%) | 29 (6.2%) | ||
| AF | 9 (4.2%) | 7 (2.7%) | 16 (3.4%) | ||
| Others | — | 3 (1.2%) | 3 (0.6%) | ||
| Axis | NL | 190 (91.3%) | 207 (81.2%) | 0.002∗ | 397 (85.7%) |
| LAD | 7 (3.4%) | 22 (8.6%) | 29 (6.3%) | ||
| RAD | 11 (5.3%) | 26 (10.2%) | 37 (8%) | ||
| P-wave | NL | 156 (76.8%) | 172 (68.3%) | 0.042∗ | 328 (72%) |
| RAE | 10 (4.9%) | 6 (2.4%) | 16 (3.5%) | ||
| LAE | 37 (18.2%) | 73 (29%) | 110 (24.2%) | ||
| BAE | — | 1 (0.4%) | 1 (0.2%) | ||
| PRI | NL | 202 (98.5%) | 246 (98%) | 0.735 | 448 (98.2%) |
| Short PRI | 2 (67%) | 3 (60%) | 5 (1.1%) | ||
| Long PRI | 1 (33%) | 2 (40%) | 3 (0.7%) | ||
| QRS | NL | 195 (97%) | 241 (95%) | 0.259 | 436 (95.8%) |
| Abnormal | 6 (3%) | 13 (5%) | 19 (4.2%) | ||
| ST | NL | 175 (84.5%) | 222 (86%) | 0.879 | 397 (85.4%) |
| STD | 9 (4.4%) | 11 (4.3%) | 20 (4.3%) | ||
| STE | 23 (11.1%) | 25 (9.7%) | 48 (10.3%) | ||
| T-wave | NL | 167 (80%) | 222 (87%) | 0.048∗ | 389 (84%) |
| Negative | 41 (20%) | 33 (13%) | 74 (16%) | ||
| Brugada | Yes | 1 (0.5%) | 3 (1%) | 0.630 | 4 (1%) |
| No | 211 (99.5%) | 253 (99%) | 464 (99%) | ||
| S1Q3T3 | Yes | 15 (7%) | 24 (9.4%) | 0.37 | 39 (8.3%) |
| No | 197 (93%) | 232 (90.6%) | 429 (91.7%) | ||
| Low voltage | Yes | 73 (34.8%) | 68 (26.4%) | 0.049∗ | 141 (30%) |
| No | 137 (65.2%) | 190 (73.6%) | 327 (70%) | ||
| ECG | NL | 46 (21.5%) | 59 (22.7%) | 0.755 | 105 (22%) |
| Abnormal | 168 (78.5%) | 201 (77.3%) | 369 (78%) | ||
| PRI | 149.06 ± 19.7 | 153.02 ± 23.2 | 0.105 | 151.2 ± 21.8 | |
| QRS | 77.7 ± 10.5 | 82.2 ± 11.4 | < 0.001∗ | 80.2 ± 11.2 | |
| QTC | 420.1 ± 39.5 | 419.5 ± 39.4 | 0.831 | 419.8 ± 39.4 | |
- ∗Statistically significant.
Table 2 shows the ECG changes of our patients in three age groups: < 45 years, 45–65 years, and > 65 years. Only 15% of our elderly patients had normal ECG, and the most common abnormalities in this group were rate and rhythm abnormality, AF, and ST changes (Table 2).
| Variable | Less than 45 (n = 122) | Between 45 and 65 (n = 151) | More than 65 (n = 158) | p value | |
|---|---|---|---|---|---|
| Rate (mean)/min | 92.8 ± 18.06 | 90.6 ± 16.8 | 85.6 ± 22.4 | 0.002∗ | |
| Rate | NL | 78 (63.4%) | 117 (77.5%) | 106 (67.1%) | 0.005∗ |
| Tachycardia | 40 (32.5%) | 32 (21.2%) | 38 (24.1%) | ||
| Bradycardia | 5 (4.1%) | 2 (1.3%) | 14 (8.9%) | ||
| Rhythm | NL | 62 (51.2%) | 104 (68.9%) | 91 (58%) | < 0.001∗ |
| Sinus tachycardia | 51 (42.1%) | 42 (27.8%) | 37 (23.6%) | ||
| Sinus bradycardia | 7 (5.8%) | 3 (2%) | 16 (10.2%) | ||
| AF | 1 (0.8%) | 2 (1.3%) | 10 (6.4%) | ||
| Others | 0 | 0 | 3 (1.9%) | ||
| Axis | NL | 107 (89.9%) | 135 (90%) | 120 (77.9%) | < 0.001∗ |
| LAD | — | 5 (3.3%) | 22 (14.3%) | ||
| RAD | 12 (10.1%) | 10 (6.7%) | 12 (7.8%) | ||
| P-wave | NL | 99 (81.8%) | 103 (70.1%) | 99 (66.9%) | 0.048∗ |
| RAE | 1 (0.8%) | 7 (4.8%) | 5 (3.4%) | ||
| LAE | 21 (17.4%) | 37 (25.2%) | 43 (29.1%) | ||
| BAE | 0 | 0 | 1 (0.7%) | ||
| PRI | NL | 118 (96.7%) | 146 (98%) | 145 (99.3%) | 0.063 |
| Short PRI | 4 (3.3%) | 1 (0.7%) | 0 | ||
| Long PRI | 0 | 2 (1.3%) | 1 (0.7%) | ||
| QRS | NL | 114 (98.3%) | 144 (98%) | 141 (91.6%) | 0.012∗ |
| Abnormal | 2 (1.8%) | 3 (2%) | 13 (8.2%) | ||
| ST | NL | 106 (88.4%) | 132 (88.6%) | 120 (78%) | 0.012∗ |
| STD | 1 (0.8%) | 6 (4%) | 13 (8.4%) | ||
| STE | 13 (10.8) | 11 (7.4%) | 21 (13.6%) | ||
| T-wave | NL | 108 (88.5%) | 127 (85.8%) | 119 (78.8%) | 0.072 |
| Negative | 14 (11.5%) | 21 (14.2%) | 32 (21.2%) | ||
| Brugada | Yes | 0 | 2 (1.3%) | 1 (0.6%) | 0.639 |
| No | 121 (100%) | 147 (98.7%) | 155 (99.4%) | ||
| S1Q3T3 | Yes | 22 (18%) | 8 (5.4%) | 8 (5%) | < 0.001∗ |
| No | 99 (82%) | 141 (94.6%) | 148 (95%) | ||
| Low voltage | Yes | 26 (22%) | 39 (26%) | 64 (41%) | 0.001∗ |
| No | 94 (78%) | 110 (74%) | 93 (59%) | ||
| ECG | NL | 29 (23.8%) | 43 (28.5%) | 24 (15.2%) | 0.017∗ |
| Abnormal | 93 (76.2%) | 108 (71.5%) | 134 (84.8%) | ||
| PRI PRI | 149.01 ± 21.6 | 149.7 ± 20.8 | 154.2 ± 22.8 | 0.176 | |
| QRSQRS | 78.3 ± 10.1 | 80.4 ± 7.9 | 82.4 ± 14.1 | 0.045 | |
| QTCQ QTC | 422.06 ± 36.8 | 416.4 ± 39.4 | 421.9 ± 42.8 | 0.308 | |
- ∗Statistically significant.
In Table 3, the ECG changes are compared based on the severity of the disease. Only 8.2% of patients who died had normal ECG and AF, STE, and tachycardia were significantly more in dead patients (Table 3).
| Variable | Nonsevere (n = 194) | Severe (n = 225) | Death (n = 49) | p value | |
|---|---|---|---|---|---|
| Rate (mean)/min | 88.3 ± 20.1 | 88.5 ± 17.1 | 98.4 ± 23.8 | 0.013∗ | |
| Rate | NL | 132 (68%) | 172 (76%) | 23 (47%) | 0.001∗ |
| Tachycardia | 50 (26%) | 46 (20%) | 23 (47%) | ||
| Bradycardia | 12 (6%) | 8 (4%) | 3 (6%) | ||
| Rhythm | NL | 115 (59.6%) | 139 (62.3%) | 22 (45%) | 0.044∗ |
| Sinus tachycardia | 56 (29%) | 67 (30%) | 19 (38.8%) | ||
| Sinus bradycardia | 15 (7.8%) | 10 (4.5%) | 3 (6%) | ||
| AF | 4 (2.1%) | 7 (3.1%) | 5 (10.2%) | ||
| Others | 3 (1.6%) | 0 | 0 | ||
| Axis | NL | 161 (85.2%) | 192 (86.5%) | 38 (82.6%) | 0.218 |
| LAD | 8 (4.2%) | 16 (7.2%) | 5 (10.9%) | ||
| RAD | 20 (10.6%) | 14 (6.3%) | 3 (6.5%) | ||
| P-wave | NL | 141 (75.4%) | 155 (71.1%) | 27 (61.4%) | 0.369 |
| RAE | 7 (3.7%) | 6 (2.2%) | 2 (4.5%) | ||
| LAE | 39 (20.9%) | 56 (25.7%) | 15 (34.1%) | ||
| BAE | 0 | 1 (0.4%) | — | ||
| PRI | NL | 187 (98.4%) | 213 (98.6%) | 42 (95.5%) | 0.074 |
| Short PRI | 3 (1.6%) | 2 (0.9%) | — | ||
| Long PRI | 0 | 1 (0.5%) | 2 (4.5%) | ||
| QRS | NL | 177 (95.2%) | 210 (96.3%) | 44 (97.8%) | 0.774 |
| Abnormal | 9 (4.8%) | 7 (3.2%) | 1 (2.2%) | ||
| ST | NL | 169 (88.5%) | 192 (86.9%) | 31 (66%) | 0.001∗ |
| STD | 7 (3.7%) | 6 (2.7%) | 6 (12.7%) | ||
| STE | 15 (7.8%) | 23 (10.4%) | 10 (21.3%) | ||
| T-wave | NL | 163 (86.2%) | 185 (84%) | 36 (75%) | 0.165 |
| Negative | 26 (13.8%) | 35 (16%) | 12 (25%) | ||
| Brugada | Yes | 2 (1%) | 2 (0.9%) | 0 | 1.000 |
| No | 190 (99%) | 220 (99.1%) | 48 (100%) | ||
| S1Q3T3 | Yes | 20 (10.5%) | 18 (8%) | 1 (2%) | 0.169 |
| No | 171 (89.5%) | 205 (92%) | 47 (98%) | ||
| Low voltage | Yes | 56 (29%) | 63 (28.5%) | 19 (39%) | 0.351 |
| No | 136 (71%) | 158 (71.5%) | 30 (61%) | ||
| ECG | NL | 47 (24.2%) | 51 (22.7%) | 4 (8.2%) | 0.036∗ |
| Abnormal | 147 (75.8%) | 174 (77.3%) | 45 (91.8%) | ||
| PRI | 151.8 ± 22.3 | 150.07 ± 20.9 | 154.6 ± 23.8 | 0.480 | |
| QRS | 80.4 ± 12.7 | 79.4 ± 8.9 | 82.2 ± 12.8 | 0.455 | |
| QTC | 417.2 ± 35.3 | 419.6 ± 39.1 | 429 ± 54.7 | 0.273 | |
- ∗Statistically significant.
Among the 474 patients, 15 required ICU care but AF “LAE” long PRI, STE, and tachycardia were significantly observed more frequently in this group (Table 4).
| Variable | ICU | p value | ||
|---|---|---|---|---|
| Yes (n = 15) | No (n = 459) | |||
| Rate (mean)/min | 97.6 ± 22.8 | 89.09 ± 19.2 | 0.108 | |
| Rate | NL | 7 (46.7%) | 325 (70.7%) | 0.057 |
| Tachycardia | 8 (53.3%) | 11 (24.1%) | ||
| Bradycardia | 0 | 24 (5.2%) | ||
| Rhythm | NL | 7 (46.7%) | 274 (60.1%) | 0.05∗ |
| Sinus tachycardia | 5 (33.3%) | 137 (30%) | ||
| Sinus bradycardia | 0 | 29 (6.4%) | ||
| AF | 3 (20%) | 13 (2.9%) | ||
| Others | 0 | 3 (0.7%) | ||
| Axis | NL | 10 (76.9%) | 387 (86%) | 0.259 |
| LAD | 2 (15.4%) | 27 (6%) | ||
| RAD | 1 (7.7%) | 36 (8%) | ||
| P-wave | NL | 4 (33.3%) | 324 (73.1%) | 0.012∗ |
| RAE | 0 | 16 (3.6%) | ||
| LAE | 8 (66.7%) | 102 (23%) | ||
| BAE | 0 | 1 (0.2%) | ||
| PRI | NL | 10 (83.3%) | 438 (98.6%) | 0.002∗ |
| Short PRI | 0 | 5 (1.1%) | ||
| Long PRI | 2 (16.7%) | 1 (0.2%) | ||
| QRS | NL | 14 (93.3%) | 422 (96%) | 0.478 |
| Abnormal | 1 (6.7%) | 18 (4%) | ||
| ST | NL | 9 (60%) | 388 (86.2%) | 0.015∗ |
| STD | 2 (13.3%) | 18 (4%) | ||
| STE | 4 (26.7%) | 44 (9.8%) | ||
| T-wave | NL | 13 (86.7%) | 376 (83.9%) | 1.000 |
| Inverted | 2 (13.3%) | 72 (16.1%) | ||
| Brugada | Yes | 0 | 4 (0.9%) | 1.000 |
| No | 15 (100%) | 449 (99.1%) | ||
| S1Q3T3 | Yes | 0 | 39 (8.6%) | 0.626 |
| No | 15 (100%) | 414 (91.4%) | ||
| Low voltage | Yes | 4 (26.7%) | 137 (30.2%) | 1.000 |
| No | 11 (73.3%) | 316 (69.8%) | ||
| ECG | NL | 2 (13.3%) | 103 (22.4%) | 0.539 |
| Abnormal | 13 (86.7%) | 356 (77.6%) | ||
| PRI | 178.3 ± 32.4 | 150.5 ± 21.02 | 0.002∗ | |
| QRS | 88.13 ± 13.7 | 79.9 ± 11.07 | 0.022∗ | |
| QTC | 445.9 ± 59.07 | 419.09 ± 38.6 | 0.150 | |
- ∗Statistically significant.
In Table 5, we summarize the ECG changes in two genders, age groups, in patients who died and in ICU, so that it is easier to compare the changes.
| Variable | Female | Male | Age < 45 years | Age ≥ 65 years | Dead | ICU | Total |
|---|---|---|---|---|---|---|---|
| Tachycardia | 54 (25.2%) | 65 (24.9%) | 40 (32.5%) | 38 (24.1%) | 23 (47%) | 8 (53.3%) | 119 (25%) |
| Dysrhythmia | 86 (40%) | 101 (39%) | 59 (49%) | 63 (40%) | 27 (54%) | 8 (53%) | 187 (40%) |
| Sinus tachycardia | 68 (31.9%) | 74 (28.7%) | 51 (42%) | 37 (23.6%) | 19 (38.8%) | 5 (33.3%) | 142 (30%) |
| AF rhythm | 9 (4.2%) | 7 (2.7%) | 1 (0.8%) | 10 (6.4%) | 5 (10.2%) | 3 (20%) | 16 (3.4%) |
| LAD | 7 (3.4%) | 22 (8.6%) | 0 | 22 (14.3%) | 5 (10.9%) | 2 (15.4%) | 29 (6.3%) |
| RAD | 11 (5.3%) | 26 (10.2%) | 12 (10.1%) | 12 (7.8%) | 3 (6.5%) | 1 (7.7%) | 37 (8%) |
| LAE | 37 (18.2%) | 73 (29%) | 21 (17.4%) | 43 (29.1%) | 15 (34.1%) | 8 (66.7%) | 110 (24%) |
| RAE | 10 (4.9%) | 6 (2.4%) | 1 (0.8%) | 5 (3.4%) | 2 (4.5%) | 0 | 16 (3.5%) |
| ORS | 13 (5%) | 6 (3%) | 2 (1.8%) | 13 (8.2%) | 1 (2.2%) | 1 (6.7%) | 19 (4.2%) |
| Repol.Abn | 62 (29%) | 57 (22%) | 26 (21%) | 49 (32%) | 21 (42%) | 6 (40%) | 119 (25%) |
| STE | 23 (11.1%) | 25 (9.7%) | 13 (10.8%) | 21 (13.6%) | 10 (21.3%) | 4 (26.7%) | 48 (10.3%) |
| STD | 9 (4.4%) | 11 (4.3%) | 1 (0.8%) | 13 (8.4%) | 6 (12.7%) | 2 (13.3%) | 20 (4.3%) |
| Invert T | 41 (20%) | 33 (13%) | 14 (11.5%) | 32 (21.2%) | 12 (25%) | 2 (13.3%) | 74 (16%) |
| Low voltage | 73 (34.8%) | 68 (26.4%) | 26 (22%) | 64 (41%) | 19 (39%) | 4 (26.7%) | 141 (30%) |
| Abnormal ECG | 168 (78.5%) | 201 (77.3%) | 93 (76.2%) | 134 (84.8%) | 45 (91.8%) | 13 (86.7%) | 369 (78%) |
Univariate logistic regression analysis showed that tachycardia, dysrhythmia, AF rhythm, and repolarization abnormalities were significantly associated with mortality. Also, multivariate logistic regression analysis showed that tachycardia and ST elevation were significantly associated with patient mortality (Table 6).
| Variable | Survived (n = 424) | Died (n = 50) | Unadjusted OR for in-hospital mortality, OR (95%CI); p value | Adjusted OR for in-hospital mortality, OR (95% CI); p value |
|---|---|---|---|---|
| Tachycardia | 96 (22.6%) | 23 (46%) | 2.9 (1.60–5.32); < 0.001∗ | 2.65 (1.2–5.8); 0.015∗ |
| Dysrhythmia | 160 (38%) | 27 (54%) | 1.9 (1.06–3.4); 0.031∗ | |
| Sinus tachycardia | 123 (29%) | 19 (38%) | 1.1 (0.64–1.9); 0.717 | |
| AF rhythm | 11 (2.6%) | 5 (10%) | 4.1 (1.38–12.45); 0.011∗ | 1.01 (0.16–6.6); 0.98 |
| LAD | 23 (5.4%) | 5 (12%) | 2.5 (0.96–6.49); 0.06 | 1.3 (0.43–4.09); 0.63 |
| RAD | 34 (8.02%) | 3 (6%) | 0.76 (0.23–2.59); 0.66 | |
| LAE | 94 (22.2%) | 15 (30%) | 1.66 (0.86–3.21); 0.13 | |
| RAE | 14 (3.3%) | 2 (4%) | 1.32 (0.29–5.98); 0.72 | |
| Repolarization abn | 98 (23%) | 21 (42%) | 2.4 (1.3–4.4); 0.005∗ | |
| STD | 14 (3.3%) | 6 (12%) | 4.1 (1.5–11.3); 0.006∗ | 2.6 (0.74–9.2); 0.14 |
| STE | 37 (8.7%) | 11 (22%) | 3.05 (1.44–6.48); 0.004∗ | 2.9 (1.2–7.2); 0.023∗ |
| T negative | 60 (14%) | 13 (26%) | 2.09 (1.05–4.17); 0.036∗ | 1.3 (0.52–3.3); 0.56 |
| Low voltage | 122 (28.8%) | 19 (38%) | 1.48 (0.809–2.73); 0.20 | |
| Abnormal ECG | 323 (76%) | 46 (92%) | 3.59 (1.26–10.23); 0.016∗ | 1.3 (0.4–4.4); 0.63 |
- ∗Statistically significant.
4. Discussion
4.1. Main Text
This retrospective study was conducted on 474 hospitalized COVID-19 positive patients to detect any electrocardiographic disorder during hospitalization.
In this study, overall, abnormal ECG was observed in 78% of patients, and was not different in women and men (78.5%V.S 77.3%), and 76.2% in patients under 45 years old but increased in more than 65 years old (84.4%), ICU-admitted (86.7%) and patients who had died (91.8%). So, more ECG changes are observed in ill patients. The results of a review study by Angeli et al. showed that about 93% of critical COVID-19 ill patients had ECG changes [8].
Mechanisms of COVID-19 cardiac damage included cytokine release syndrome, direct myocardial injury by interaction between angiotensin-converting enzyme 2 (ACE2) and virus [9], coronary spasm, plaque instability, hypercoagulable state, acute coronary syndrome, and cardiac toxicity by antiviral medicines, steroids, and electrolyte abnormalities [15]. This process can damage the myocardium directly (cytokine storm or viral injury) or indirectly through inflammation, thrombosis, endothelial dysfunction, hypoxia, and stress-induced arrhythmias, both of which (direct and indirect cardiac damage) causes a variety of ECG changes [10, 16].
In our study, the most common ECG abnormality in COVID-19 patients was dysrhythmia (40%), which did not differ in women and men, but it was significantly more in ICU admitted (53%) and patients who had died (54%). Dysrhythmias related to COVID-19 can be due to a direct effect on the heart, such as myocarditis, ischemia, pericarditis, and pulmonary emboli, or secondary to the COVID-19 effect on other organs, such as hypoxemia due to sepsis or respiratory distress or electrolyte imbalance related to kidney injury [11].
The most reported dysrhythmia in this study includes sinus tachycardia (30.1%), sinus bradycardia (6.2%), and AF (3.4%). Sinus tachycardia (23.6%) and AF (6.4%) were more in ≥ 65 years old. AF is directly related to the disease severity, as 10.2% of patients died and 20% of admitted ICU patients had AF. AF is related to inflammatory conditions such as cardiomyopathy from COVID-19, which happens in 50% of patients hospitalized in ICU [14, 17, 18].
Sinus tachycardia and AF independently are a sign of myocardial damage, disease severity, and poor prognosis in COVID-19 patients [12]. Other mechanisms such as reduced availability of ACE, binding of viral spike protein to CD147, or sialic acid have been proposed as the pathophysiology of AF associated with COVID-19. These possible mechanisms cause endothelial damage, cytokine storm, and increased adrenergic levels [19]. Pinto-Filho, in a multicenter study on 2451 COVID-19 patient, reported that sinus tachycardia, especially heart rate higher than 120 per minute, had the strongest relationship with mortality (OR = 3.4) [20].
Other tachyarrhythmias such as atrial tachycardia, ventricular tachycardia, and Torsade de Pointes were not reported in our study. The reason for the difference between the results of our study and other studies can be the different patient populations and treatment protocols.
Sinus bradycardia was present in 6.2% of total patients and 10% of our elderly patients. Sinus bradycardia in COVID-19 patients can be due to the sinoatrial (SA) node or atrioventricular (AV) node damage by cytokines releasing acute myocarditis, severe hypoxia, and antiviral medicine like azithromycin, hydroxychloroquine, lupinavir, and remdesivir side effects [21].
Tachycardia (HR ≥ 100/min with any rhythm) was observed in 25% of our patients; it was directly related to the severity of the disease, as 47% of dead patients and 53.3% of admitted patients in ICU had tachycardia (OR = 2.65, CI 1.2–5.8, p = 0.015). The results are similar to the Lanza study, where more than a third of patients had tachycardia [21]. Tachycardia can be a response to increased cardiac demand during the infection, fever pain, hypovolemia, hypoperfusion, hypoxia, and anxiety [12].
Repolarization abnormality including ST elevation, ST depression, and inverted wave was observed in 25% of the hospitalized COVID-19 patients and 40% of ICU admitted patients in our study, with multiple logistic regression; ST elevation was associated with disease severity with OR 2.9 (CI 1.2–7.2; p = 0.023).
ST segment change and inverted T wave are one of the most challenging changes in ECG during the COVID-19 pandemic. ST-T changes were related to myocardial ischemia. Myocardial ischemia can be caused by an imbalance between supply and demand, which is more common in critically ill patients or coronary artery stenosis. In line with our study, Li et al. observed ST-T changes in 40% of patients, which was significantly higher in patients hospitalized in ICU (65.2% vs. 34.5%) than in patients not hospitalized in ICU [22]. The results of Lanza’s study also showed that ST elevation has a strong correlation with mortality in COVID-19 patients [21].
Antwi-Amoabeng et al. reported in a study on 186 patients of COVID-19 that 8.1% of patients had ST elevation, but ST depression had a strong relationship with mortality, so that ST depression was observed in 8.6% of patients, in 4.5% alive, and 28.1% of dead people [23].
In our study, low voltage was observed in 30% of patients, which was significantly higher in women (34.8% vs. 26.4%) and people over 65 years old (41% vs. 22%), but it was not related to mortality. In a study in Iran, Mirtajaddini reported low voltage in ECG in 54.5% of patients with COVID-19. Of course, all these patients had ST elevation in the ECG. No correlation between low voltage and mortality or heart failure was reported in their study [24]. However, in a study on 1258 hospitalized COVID-19 patients, only 3.4% had low voltage in the ECG, which was not related to mortality or hospitalization in the ICU [25]. We did not find a reason for the low voltage in women and elderly people. In a recent study, Tso et al. reported that the low voltage in the ECG was much more common in female athletes than in male athletes [26].
In our study, RAD and LAD axis deviations were observed in 8% and 6.3% of patients, respectively. In COVID-19, massive pulmonary embolism or acute respiratory failure causes right ventricular strain. RV strain causes changes in the patient’s ECG, such as right axis deviation, T wave inversion, and ST depression in inferior and anteroseptal leads (V1–V4), which are also prominent R waves in leads V1 and V2 [27].
4.2. Limitation and Strength
Our study had some limitations. The most important of them is the retrospective study, which cannot investigate the causality. Also, we did not have the old EKGs of the patients to check whether the changes were new or old. Since the ECG changes can be due to the effects of the drugs used to treat COVID-19, we did not examine the patients′ drugs in this study. One of the other limitations of our study is that we could not assess all hospitalized patients during the study period because some patients did not have an ECG.
4.3. Conclusions
Our study shows that most of the COVID-19 patients hospitalized, especially people admitted to ICU or dead, had ECG abnormalities. Dysrhythmia and repolarization disorder were the most common ECG changes. ST elevation and tachycardia (rate above 100) are strongly related to disease severity and mortality. Therefore, it seems necessary to take electrocardiography in patients with viral infections such as COVID-19, which is a cheap, available, and noninvasive diagnostic method to check the cardiac function easily, especially in patients with more severe disease or hospitalized in ICU. By the way, in our study, the amount of low voltage was relatively high, which needs further investigation. We suggest that more studies should be done, especially investigating the impact of different COVID-19 treatments on ECG changes.
Nomenclature
-
- ECG
-
- Electrocardiogram
-
- ICU
-
- Intensive care unit
-
- Tachy
-
- Tachycardia
-
- Brady
-
- Bradycardia
-
- AF
-
- Atrial fibrillation
-
- LAD
-
- Left axis deviation
-
- RAD
-
- Right axis deviation
-
- LAE
-
- Left atrial enlargement
-
- RAE
-
- Right atrial enlargement
-
- BAE
-
- BI atrial enlargement
-
- PRI
-
- PR interval
-
- SA
-
- Sinoatrial
-
- AV
-
- Atrioventricular
-
- AVN
-
- Atrioventricular node
-
- ACE
-
- Angiotensin-converting enzyme
Ethics Statement
The ethical code was obtained (IR.BUMS.REC.1399.546), and patients signed informed consent.
Conflicts of Interest
The authors declare no conflicts of interest.
Author Contributions
Study conception and design: Toba Kazemi, Seyed Mohammad Riahi, Abdol Sattar Pagheh, and Mohammad Reza Khazdair.
Data collection: Mohamad Karimi, Mohammad Yousof Qoddusi, Moloud Foogerdi, Anahita Arian, Shima Heidary, and Toba Kazemi.
Supervision: Toba Kazemi, Abdol Sattar Pagheh, and Mohammad Reza Khazdair.
Data analysis: Fateme Mahdizadeh and Seyed Mohammad Riahi.
Interpretation of results: Sima Sobhani Shahri, Fateme Mahdizadeh, Seyed Mohammad Riahi, and Toba Kazemi.
Draft manuscript preparation: Sima Sobhani Shahri, Fateme Mahdizadeh, and Toba Kazemi.
Final approval of manuscript: All authors.
Funding
This article was the result of a research project conducted by faculty members of Birjand University of Medical Sciences with code 5683. Part of the budget for this project was provided by the Deputy for Research and Technology of the Ministry of Health and Medical Education of Iran and part of it was provided by the Deputy for Research and Technology of Birjand University of Medical Sciences.
Acknowledgments
The authors would like to thank the Clinical Research Development Unit of Vali-Asr Hospital, Birjand University of Medical Sciences. This article was the result of a research project conducted by faculty members of Birjand University of Medical Sciences with code 5683. Part of the budget for this project was provided by the Deputy for Research and Technology of the Ministry of Health and Medical Education of Iran and part of it was provided by the Deputy for Research and Technology of Birjand University of Medical Sciences.
Open Research
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author.
