Introduction

The coronavirus 2019 (COVID-19), first identified in Wuhan in December 2019, primarily affects the respiratory system.¹,² However, COVID-19 is also known to affect the other system in the body and can lead to extrapulmonary involvements, including cardiac arrhythmia.³,⁴ An early study from Wuhan reported an incidence of arrhythmia in 16.7% of hospitalized COVID-19 patients.⁵ In addition, the incidence was higher in patients admitted to the Intensive Care Unit (ICU) and reached 44%.⁵ However, the study did not describe the specific type of arrhythmia.⁵ The severity of the disease was also found to correlate with the incidence of arrhythmia, with a higher incidence in patients with severe disease compared to mild disease.⁶

In some COVID-19 patients, arrhythmia may be the first infection symptom, preceding the respiratory symptoms.⁶ COVID-19-related arrhythmia can manifest as tachyarrhythmia or bradyarrhythmia.⁷-¹¹ Atrial fibrillation (AF) was the most commonly reported supraventricular tachyarrhythmia.¹²,¹³ In addition, ventricular tachyarrhythmia such as ventricular tachycardia (VT) was reported in hospitalized COVID-19 patients.⁸,¹⁴,¹⁵ Several studies also reported bradyarrhythmia in the form of sinus bradycardia, total atrioventricular (AV) block, and sinus pause.⁸,¹⁰,¹¹,¹⁴,¹⁵

Despite accumulating evidence, there is still variation in the type and outcomes of arrhythmia associated with COVID-19 in hospitalized patients. Some retrospective studies did not include COVID-19 patients with asymptomatic and mild disease severity. Therefore, this study aims to investigate and to compare the outcomes of hospitalized COVID-19 patients with and without arrhythmia from asymptomatic to life-threatening disease severity and to investigate the predictors that may influence the in-hospital mortality, the need for intensive care (IC), the need for intubation and mechanical ventilation (MV), hypotension requiring vasopressor, thromboembolic event (deep vein thrombosis and pulmonary embolism), major bleeding, and stroke or transient ischemic attack (TIA).

Material and methods

Inclusion and exclusion criteria. This study included moderate, severe, and critically ill COVID-19 patients ≥ 18 years who were hospitalized with the primary indications of COVID-19 and tested positive with real time-polymerase chain reaction (RT-PCR) between April 1, 2021, and December 31, 2021, at Gunung Jati General Hospital, Cirebon. We also included patients admitted to our hospital due to other medical indications with positive RT-PCR tests. The COVID-19 severity of patients admitted due to other medical indications will be assessed and categorized as asymptomatic, mild, moderate, severe, or critically ill. Patients who were pregnant, discharged, or transferred by their intention before fulfilling treatment were excluded.

Data collection. Data on patients’ demographic, clinical characteristics, comorbidities, history of cardiovascular procedures, electrocardiography, management during hospitalization, outcomes were directly collected from hospital medical records. Two cardiologists interpreted the electrocardiography. Any disagreement in the electrocardiography interpretation was resolved through discussion. Continuous telemetry monitoring was performed on all patients admitted to the ICU. We included major arrhythmias (supraventricular tachycardia, new-onset atrial fibrillation, new-onset atrial flutter, pre-existing atrial fibrillation, pre-existing atrial flutter, atrioventricular block, sinus pause, sinus arrest, ventricular tachycardia, ventricular fibrillation) and non-major arrhythmias (sinus tachycardia, sinus bradycardia, premature atrial complex, premature ventricular complex).⁷,⁸ New-onset atrial fibrillation (NOAF) or atrial flutter is defined as electrocardiographic evidence of atrial fibrillation or atrial flutter during admission or hospitalization in patients without a medical history of atrial fibrillation or atrial flutter. Pre-existing atrial fibrillation (PEAF) or atrial flutter is defined as patients with a medical history of atrial fibrillation or atrial flutter. The outcomes were in-hospital mortality, the need for IC, the need for intubation and mechanical ventilation, hypotension requiring vasopressor, thromboembolic event (deep vein thrombosis and pulmonary embolism), major bleeding, and stroke/TIA. Our hospital’s ethics committee approved this study with the number No. 087/LAIKETIK/KEPKRSGJ/VI/2021.

Statistical analysis. Descriptive statistics were used to summarize the data. Categorical variables were presented as frequencies and percentages. The Shapiro-Wilk or Kolmogorov-Smirnov normality test was performed as appropriate. Normally distributed data for continuous variables were summarized as mean and standard deviation. Otherwise, the data were presented as the median and interquartile range (25^th percentile = P25, 75^th percentile = P75). The independent T-test or Mann-Whitney U-test was performed to compare continuous variables between patients with and without arrhythmia. Chi-square (χ²) and Fisher’s exact were used to compare dichotomous variables between patients with and without arrhythmia. Results were considered significant at p-value < 0.05.

Furthermore, the χ² test or Fischer’s exact test was performed to compare and determine the independent variables according to the outcomes of patients with arrhythmia that will be included in the multivariate logistic regression analysis. Variables with χ² or Fisher’s exact p-value logistic regression and backward method were applied. The statistical analysis was conducted using SPSS version 22.0.

Results

This cross-sectional studies included 257 patients admitted due to confirmed COVID-19, from April 1 to December 31, 2021, for study analysis. A total of 94 patients (36.6%) had arrhythmia during hospitalization. The patients’ demographic data are shown in Table 1. Most patients were male (n = 137, 53.3%), and the mean age was 53.49 ± 13.34 years old. The proportion of IHD in patients with arrhythmia was almost twice that of patients without arrhythmia, although the difference did not reach statistical significance.

As seen in Table 2, patients with arrhythmia had more severe disease (48.9% vs. 22.7%; p < 0.001) than those without arrhythmia, as reflected in clinical presentation and laboratory results. Compared to patients without arrhythmia, patients with arrhythmia had lower oxygen saturation during initial admission (90.5 vs. 96; p < 0.001). Patients with arrhythmia had a higher level of leukocytes (9,465 vs. 8,070/µL; p = 0.019), neutrophil-to-lymphocyte ratio (NLR) (6.53 vs. 4.62; p < 0.001), creatinine (1.34 vs. 1.14 mg/dL; p = 0.022), and d-dimer level (2,061 vs. 1,167 ng/mL; p < 0.001) compared to patients without arrhythmia, as presented in Table 3. The high-flow nasal cannula [23 (24.5%) vs. 15 (9.2%); p = 0.001], anticoagulant [80 (85.1%) vs. 117 (71.8%); p = 0.015], and plasma convalescent [28 (29.8%) vs. 27 (16.6%); p = 0.013] were more commonly prescribed in patients with arrhythmia compared to those without arrhythmia, as seen in Table 4.

In this study, 94 patients (36.6%) displayed arrhythmia during hospitalization. The electrocardiographic parameter between patients with and without arrhythmia can be seen in Table 5; the arrhythmias are listed in Table 6. The patients can present with single or multiple arrhythmias during hospital admission. From patients with single arrhythmia, the most common arrhythmia was sinus tachycardia (69.1%) and followed by pre-existing AF (10.6%). New-onset AF occurred in 2.1% of patients during the hospitalization period. Total AV block also occurred in 2.1% of patients. For patients with multiple arrhythmias, the most common arrhythmia was sinus tachycardia with premature atrial complex (3.2%), followed by sinus tachycardia with premature ventricular complex (1.1%) and sinus bradycardia with first degree AV block and premature ventricular complex (1.1%).

Overall, patients with arrhythmia had significantly worse outcomes than those without during hospitalization, as shown in Table 7. The need for intubation and mechanical ventilation [7 (7.4%) vs. 2 (1.2%); p = 0.009] and hypotension requiring vasopressor [15 (16%) vs. 5 (3.1%); p < 0.001] were more commonly observed in patients with arrhythmia compared to patients without arrhythmia. Due to higher disease severity and complications, the need for IC was significantly higher in patients with arrhythmia compared to those without arrhythmia [47 (50%) vs. 38 (23.3%); p < 0.001].

Age ≥ 65 years old (OR 3.203; 95% CI 1.117-9.186; p = 0.026), DM (OR 2.918; 95% CI 1.248-6.824; p = 0.012), leukocytosis during admission (OR 2.616; 95% CI 1.134-6.033; p = 0.023), and NLR > 6.82 during admission (OR 2.641; 95% CI 1.139-6.123; p = 0.017) were associated with the need for IC in univariate analysis. After multivariate analysis, only DM (adjusted OR 2.656; 95% CI 1.114-6.333; p = 0.028) remained associated with the need for IC and can be seen in Table 8.

In univariate analysis, history of stroke/TIA (OR 17; 95% CI 1.965-147.046; p = 0.027) and NLR > 6.82 (OR 8.914; 95% CI 1.028-77.292; p = 0.041) were associated with the need for intubation and mechanical ventilation as seen in Table 8. After multivariate analysis, history of stroke/TIA (adjusted OR 46.426; 95% CI 1.64-1314.392; p = 0.024), thrombocytopenia during admission (adjusted OR 17.131; 95% CI 1.623-180.777; p = 0.018), and NLR > 6.82 (adjusted OR 0.67; 95% CI 0.005-0.882; p = 0.04) associated with the need for intubation and mechanical ventilation. In univariate analysis, new-onset AF and premature atrial complex (PAC) numerically increased the need for intubation and mechanical ventilation.

In univariate analysis, age ≥ 65 years old (OR 5.802; 95% CI 1.79-18.805; p = 0.004), DM (OR 7.286; 95% CI 1.895-28.007; p = 0.001), CKD (OR 3.9; 95% CI 1.021-14.892; p = 0.036), leukocytosis during admission (OR 4.039; 95% CI 1.181-13.810; p = 0.019), and NLR > 6.82 during admission (OR 4.492; 95% CI 1.311-15.387; p = 0.011) were associated with hypotension requiring vasopressor. After multivariate analysis, only DM (adjusted OR 4.850; 95% CI 1.172-20.078; p = 0.029) remained associated with hypotension requiring vasopressor, as shown in Table 8.

Regarding in-hospital mortality, age ≥ 65 years old (OR 5.785; 95% CI 1.901-17.598; p = 0.001), DM (OR 8.315; 95% CI 3.270-21.141; p < 0.001), CKD (OR 2.813; 95% CI 1.199-6.6; p = 0.016), leukocytosis during admission (OR 2.778; 95% CI 1.199-6.346; p = 0.016), and NLR > 6.82 during admission (OR 2.745; 95% CI 1.183-6.371; p = 0.017) were associated with in-hospital mortality in univariate analysis. Atrial arrhythmias were also found to increase the risk of in-hospital mortality. After multivariate analysis, only DM (adjusted OR 6.52; 95% CI 2.445-17.387; p < 0.001) remained associated with in-hospital mortality, as shown in Table 8.

Discussion

This study involved 257 patients with confirmed COVID-19. The prevalence of arrhythmia in our study was 36.6%, which was higher than the previously reported prevalence.⁵,⁶,¹⁶,¹⁷ We found that the most common arrhythmia was sinus tachycardia. This type of arrhythmia was in line with the previous study.⁷ No VT or ventricular fibrillation was detected in our study. The need for IC, intubation and MV, hypotension requiring vasopressor, and in-hospital mortality were significantly higher in patients with arrhythmia than those without arrhythmia.

The mean age of arrhythmia patients was 52.48 ± 13.94 years old. The mean age from our study was younger compared with earlier studies.⁵-⁷,¹⁴ The younger study population may be explained by the admission criteria in our hospital. Our hospital is one of the referral COVID-19 hospitals in West Java. Therefore, we included patients transferred from other hospitals and primary healthcare facilities and traced from communities.

Based on previous studies, arrhythmia was associated with a critical illness.⁶,¹⁶ From our findings, severe and critical disease severity was more commonly observed in patients with arrhythmia compared to patients without arrhythmia. Higher proportions of severe and critical COVID-19 cases may contribute to elevated in-hospital mortality among hospitalized patients with arrhythmia. Disease severity was reflected in vital signs, comorbidities, and laboratory results in patients with arrhythmia. The oxygen saturation was significantly lower in patients with arrhythmia during initial admission. Low presenting oxygen saturation was associated with severe disease.¹⁸ Previous studies reported that several comorbidities, such as hypertension, DM, HF, and previous arrhythmia, were more prevalent in patients with arrhythmia.⁶,¹⁶ Indeed, patients with arrhythmia in our study had significantly higher comorbidities (e.g. hypertension, HF, DM, CKD) than those without arrhythmia. The presence of comorbidities, including PEAF, increased the susceptibility to develop a severe COVID-19 disease.¹⁹ Moreover, the neutrophil count, NLR, and D-dimer were significantly elevated compared to patients without arrhythmia. These inflammatory markers were significantly elevated in severe or critical COVID-19 cases.²⁰ Furthermore, the level of C-reactive protein (CRP), one of the inflammatory markers, was higher in patients with arrhythmia than in patients without, although the difference was not statistically significant.

Based on our cohorts, severe COVID-19 disease was more prevalent in patients with arrhythmia than those without, potentially increasing the risk of poor hospital outcomes for patients with arrhythmia. Theoretically, the combination of low oxygen saturation, high levels of inflammatory markers, and high level of proinflammatory cytokine during severe COVID-19 disease may lead to clinical deterioration during hospitalization.²¹-²³ In addition, arrhythmia-related changes in heart rate, whether too fast or too slow, may cause clinical instability.¹¹,²⁴,²⁵ However, our univariate and multivariate analysis did not find arrhythmia to be a predictor of in-hospital outcomes, as shown in Tables 8 to 12. Moreover, the most common arrhythmia in this study was sinus tachycardia, a minor arrhythmia with multifactorial aetiologies. Sinus tachycardia may reflect the more severe clinical features of COVID-19 disease but is less likely to cause clinical deterioration. On the other hand, NOAF and AV block may be more significant predictors of in-hospital outcomes than sinus tachycardia. However, neither NOAF nor AV block achieved a p-value < 0.25 and were not included as variables in our multivariate analysis. Thus, it appears that arrhythmia in hospitalized COVID-19 patients may be a marker of the disease’s severity rather than a variable that worsens the COVID-19 disease. In addition, univariate analysis was performed to identify the association between atrial fibrillation type and in-hospital outcomes, as presented in Tables 9 to 12. Neither NOAF nor PEAF were associated with in-hospital outcomes. However, NOAF is more easily attributable to the COVID-19 infection itself compared to PEAF. Suggesting the NOAF may act as a risk marker of COVID-19 infection, while PEAF may act as a risk factor to worse COVID-19 infection during hospitalization period.

From our study, the need for IC, hypotension requiring vasopressor, and in-hospital mortality were higher in patients with arrhythmia. Our study found an in-hospital mortality rate as high as 44.7%, and this in-hospital mortality rate of patients with arrhythmia was higher than in the previous studies.⁶-⁸,¹⁶ Multivariate analysis showed that DM in patients with arrhythmia was associated with the need for ICU, hypotension requiring vasopressor, and in-hospital mortality. Diabetes mellitus can increase disease severity and progression to cardiorespiratory failure by increasing inflammatory cytokines, natural killer cells, reactive oxygen species, interleukin-6, D-dimer, and fibrinogen levels.²⁶ Diabetes mellitus in COVID-19 was also associated with a two-fold increase in mortality compared to those without DM.²⁷ Moreover, the presence of DM in our patients with arrhythmia was higher than in previous studies.⁶-⁸,¹⁶ During the period of data collection, the shortage of doctors, nurses, personal protective equipment, ventilators, and COVID-19 drugs occurred in our hospital. Furthermore, Indonesia ranked first with the most COVID-19 cases in Southeast Asia, according to the Center for Strategic and International Studies (CSIS).²⁸ Combined, this may contribute to higher in-hospital mortality rates compared to previous studies from developed countries and may also explain the higher need for IC and hypotension requiring vasopressor in patients with arrhythmia.

History of stroke, thrombocytopenia during admission, and NLR > 6.82 during admission were associated with the need for intubation and mechanical ventilation. However, the confidence interval was too wide to draw firm conclusions, and NLR > 6.82 during admission reduced the risk for intubation and mechanical ventilation after multivariate analysis. This finding was contrary to the previously reported study.²⁹ A relatively small number of patients with arrhythmia who required intubation and mechanical ventilation may explain this difference.

From our study, we reported zero incidences of thromboembolic events. At the time of writing this article, our hospital could not perform urgent Doppler ultrasonography for confirming deep vein thrombosis and urgent computed tomography pulmonary angiography or invasive pulmonary angiography for confirming pulmonary embolism. Therefore, the true incidence of either deep vein thrombosis or pulmonary embolism may be missed.

The main limitation of our study was the inability to perform continuous electrocardiographic monitoring on patients outside the ICU ward. Hence, the arrhythmia prevalence in patients with asymptomatic and mild disease may be missed. Second, the study design was cross-sectional, and data were retrospectively collected based on medical records. Changing the study design to a prospective one and contemplating the inclusion of patients with known pre-existing arrhythmia only when they have been diagnosed with COVID-19 could be a valuable consideration. Third, we failed to establish a causal link or a comparative effect between primary arrhythmia and in-hospital outcomes. Fourth, we included patients admitted due to other medical indications that tested positive for COVID-19. Fifth, the relatively small sample size of patients with arrhythmia was included in the multivariate logistic regression analysis resulting in a wide confidence interval. Sixth, some laboratory data, such as electrolyte, CRP, and D-dimer, were not complete and were not included as variables in the multivariate logistic regression analysis. Finally, markers of myocardial injury and pro-inflammatory cytokines were not measured, preventing us from assessing their association with arrhythmia in hospitalized COVID-19 patients.

Conclusions

According to our data, the in-hospital outcomes in patients with COVID-19 and arrhythmia are the worst. Severe and critical COVID-19 disease was more commonly observed in patients with arrhythmia than in patients without arrhythmia. The arrhythmia may be a marker of severe COVID-19 disease rather than a variable that worsens COVID-19 disease. Diabetes mellitus is associated with a higher need for IC, hypotension requiring vasopressor, and in-hospital mortality in patients with arrhythmia. A history of stroke/TIA and thrombocytopenia during admission are associated with a higher need for intubation and mechanical ventilation. Further prospective trials with a larger sample size are needed to confirm this finding.