Worldwide Survey of COVID-19-Associated Arrhythmias.

[Figure: see text].

Coronavirus disease 2019 (COVID-19) infection has become a leading cause of hospitalization and death worldwide, and cardiac arrhythmias have been noted in these patients.

Preexisting comorbidities, such as hypertension, diabetes, heart failure, and coronary artery disease, are common in hospitalized patients with COVID-19.

Of 827 hospitalized patients with COVID-19 who developed a cardiac arrhythmia, most (7 in 10) developed atrial arrhythmias. One in 5 hospitalized patients with COVID-19 developed ventricular arrhythmias and a similar proportion developed bradyarrhythmias.

The presence of arrhythmia was associated with significant morbidity and mortality; about 4 in 10 patients who developed an arrhythmia were mechanically ventilated and only one half survived to hospital discharge.

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), the virus that causes coronavirus disease 2019 (COVID-19), has affected over 50 million people worldwide since late December 2019, leading to over 1 million deaths.[1] COVID-19 is known to affect nearly every organ system and the heart is no exception.[2] The Heart Rhythm Society, American College of Cardiology, and the American Heart Association released a joint statement with recommendations regarding exposure risks, triage, cardiac arrhythmias as well as how to manage electrophysiology procedures, clinic visits, and device interrogations.[3]

Although the underlying pathophysiology has remained elusive, various single-center studies and surveys around the world have reported a spectrum of electrophysiology issues associated with the disease and its therapies, specifically atrioventricular heart block, atrial fibrillation (AF), and polymorphic ventricular tachycardia (VT).[4-7] Coexisting hypoxia, electrolyte disarray, and the administration of arrhythmogenic medications (eg, hydroxychloroquine, azithromycin) make it difficult to ascertain the direct and indirect contribution of COVID-19 on cardiac arrhythmias. A 58% increase in out-of-hospital cardiac arrest in the Lombardy region of Italy during the first 40 days of the COVID-19 pandemic raised concerns regarding the risk of arrhythmia associated with SARS-CoV-2 infection.[8] In the United States, a large single-center study of 700 patients demonstrated that admission to the intensive care unit was independently associated with a 10-fold increase in arrhythmia risk.[9] Although cardiac arrest was associated with in-hospital mortality, most were due to nonshockable rhythms, and incident AF, bradycardia, and nonsustained VT were not associated with mortality.[8]

In this global case series, including 29 institutions across 4 continents, 12 countries, and 28 cities (Figure, Table 1), we aimed to survey the experiences of institutions from multiple continents to better understand the type of arrhythmias associated with COVID-19 infection, as well as global differences in modulations of routine electrophysiology practice due to the pandemic. Patient care of COVID-19 infection has varied across the globe due to differences in rhythm manifestations and management of the disease but also due to differences in non-COVID-19 medical care and resources available to physicians. The incidence, diagnosis, and management of arrhythmias and conduction system disease have varied across populations, cities, countries, and continents worldwide. We will also describe how COVID-19 affected routine electrophysiology practice and procedural volume.

Table 1.

Worldwide Cases Contributed to This Case Series

Map of cases contributed to this worldwide survey.

Methods

We conducted a retrospective, international, multicenter analysis of patients ≥18 years of age with a clinical diagnosis of SARS-CoV-2 (confirmed by nasopharyngeal PCR testing) with and without incident cardiac arrhythmias while hospitalized with COVID-19 infection. Patients were classified as having arrhythmia if they had documented AF, atrial flutter, supraventricular tachycardia (SVT), nonsustained VT, sustained monomorphic or polymorphic VT, sustained ventricular fibrillation (VF), atrioventricular block, or marked sinus bradycardia (heart rate of <40 bpm). Deidentified clinical data including demographics, comorbidities, baseline electrocardiographic and echocardiographic findings, antiviral therapy, use of hydroxychloroquine with or without azithromycin, and presence of arrhythmia during hospitalization was provided by each treating institution. Patient admissions ranged from January 4, 2020 (Modena and Padova, Italy) to August 7, 2020 (Porto Alegre, Brazil), and varied based upon the timing of peak infection rate in each region. Information was also collected regarding the volume of electrophysiology procedures at each center during the pandemic. The data that support the findings published in this study are available from the corresponding author on request.

Cardiac electrophysiologists at medical centers throughout the world were contacted and asked to participate in a survey of cardiac arrhythmias occurring in patients hospitalized with COVID-19 infection. Patient-level data were collected and reported by each collaborating center utilizing a prespecified excel spreadsheet; the categories collected are included in Table I in the Data Supplement.

Columbia University Medical Center served as the coordinating center. Study data were collected and managed using REDCap electronic data capture tools hosted at Columbia University.[10,11] The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the Columbia University Institutional Review Board. The study was approved by each participating center’s respective Institutional Review Board.

Descriptive statistics were computed. Values for continuous variables are presented as mean±SD or median [interquartile range] as appropriate. Categorical variables were presented as counts and percentage. Where appropriate, Pearson χ2 test and ANOVA were used to compare proportions and group means, respectively. Statistical analysis was performed using SPSS Version V27.

Results

COVID-19 Across the Globe

Data were submitted for 4526 patients hospitalized with COVID-19 infection across the world for whom data was available, and we further analyzed 827 patients who had a cardiac arrhythmia as a result of COVID-19 infection. Overall, the mean age of the total cohort with and without arrhythmias was 62.8±17.0 years, and 57% were men (Table 2). The prevalence of medical comorbidities was high (eg, 55% had hypertension and 35% had diabetes mellitus); however, among the 4 continents surveyed, patients from Asia were generally healthier, as only 29% and 16% reported a history of hypertension and diabetes, respectively. Likewise, compared with Europe, North America, and South America, patients from Asia had lower body mass indices (23.7±4.6 kg/m2 versus mean body mass index >27 kg/m2 in all other continents represented). As expected, the majority of patients from Europe, North America, and South America identified as White or Hispanic, whereas 97% of the population from Asia identified as Asian. Treatment regimens for COVID-19 varied considerably with over 50% of patients in Europe and Asia receiving antiviral agents (oseltamivir or remdesivir) compared with only 6% to 7% of patients in North and South America. Additionally, over 60% of patients in Europe and North America received hydroxychloroquine while fewer than 20% did in Asia and South America. Ten percent of patients worldwide were treated with IL (interleukin)-6 receptor inhibitors, such as tocilizumab or sarilumab, and these medications were used rarely in Asia and South America.

Table 2.

Clinical Characteristics of Patients With COVID-19 With and Without Arrhythmia Worldwide

The majority of sites contributing did not send information on all patients admitted to their hospital with COVID-19, and thus incidence of arrhythmia cannot be estimated on the full cohort. Four institutions (International Medical University in Kuala Lumpur, Malaysia; NY-Presbyterian–Columbia and NY-Presbyterian–Queens in New York; and Hospital de Clínicas de Porto Alegre) contributed data on all patients admitted to their institutions. Among 2762 patients admitted to these institutions with COVID-19 infection, the incidence of cardiac arrhythmia AF, atrial flutter, SVT, nonsustained VT, VT, VF, marked sinus bradycardia, atrioventricular block, or pauses >3 seconds was 12.9%.

Eight hundred twenty-seven patients across all institutions developed cardiac arrhythmia while hospitalized for COVID-19 infection (Table 3). Among this cohort, the mean age was 71.1±14.1 years, and 65% were men. Cardiac comorbidities were common, as 69% had hypertension, 42% had diabetes mellitus, 30% had congestive heart failure, and 24% had coronary artery disease. The majority of patients with SARS-CoV-2 infection who developed a cardiac arrhythmia had no prior history of arrhythmia. Similar to what has been previously reported, cardiac arrhythmias were associated with significant morbidity and mortality[6]; 43% of patients who developed arrhythmia were mechanically ventilated, and only 51% survived to hospital discharge. Of the 359 patients who were mechanically ventilated, 146 (41%) survived to discharge. Similar regional variations in medical management were noted in patients with arrhythmia as in the entire cohort.

Table 3.

Clinical Characteristics of Patients With COVID-19 With Cardiac Arrhythmia Worldwide

Due to evolving understanding of the disease process and efficacy of medical therapies, we compared treatments and outcomes among patients admitted from January through April and May through August. Although the majority of patients (65%) were treated with hydroxychloroquine early in the pandemic, only 13% of patients admitted after April received this therapy. Azithromycin was used as therapy for 47% of patients early in the pandemic, compared with 31% later in the pandemic. There was no similar decrease in the use of antiviral medications over time, with 21% of patients receiving this therapy in the early period of the pandemic, compared with 17% later in the pandemic. Despite increasing experience treating the disease, there was no significant change in survival or number of patients ventilated seen in this cohort, with 49% of patients admitted January through April surviving to discharge, compared with 52% of patients admitted after April 2020. These trends were similar both in the cohort of patients with arrhythmia and the full cohort of patients with and without cardiac arrhythmia.

Of patients who developed an arrhythmia, the majority (81.8%) developed atrial arrhythmias including AF, atrial flutter, or another SVT (Table 4). However, these arrhythmias occurred less often among patients in Asia, where only 34% of patients had AF and a larger proportion of patients had bradyarrhythmia or atrioventricular block. The incidence of ventricular arrhythmias was less common than other dysrhythmias across all regions, with 21% of patients having nonsustained VT, VT, or VF. Of patients with VT, there was an equal distribution of monomorphic and polymorphic morphology. 22.6% of patients with COVID-19 infection developed bradyarrhythmias.

Table 4.

Types of Cardiac Arrhythmias in Patients With COVID-19 Worldwide

Table 5 compares the clinical characteristics and medical therapies among patients by type of arrhythmia. Ventricular arrhythmia was associated with significant mortality, and only 38% of these patients survived to discharge. Rates of mechanical ventilation and mortality were high regardless of type of arrhythmia, though patients who developed VT were 1.3× more likely to be mechanically ventilated and 1.4× more likely to die during the hospitalization than patients with atrial or bradyarrhythmias.

Table 5.

Clinical Characteristics and Arrhythmia Management of Patients With COVID-19 With Arrhythmia

Forty-five percent of patients had their left ventricular (LV) function assessed with an echocardiogram in the 2 years before admission, with a mean LV ejection fraction of 51.7±13%. Among 827 patients who developed cardiac arrhythmia, only 32% had echocardiographic assessments during their hospitalization; this may be due to substantial resource limitation, in particular at the height of the pandemic. Of those patients who underwent assessment of LV assessment during their hospitalization, the mean LV ejection fraction was 52.5±13%. At the time of hospital admission with COVID-19 infection, the corrected QT interval (Bazett) was borderline prolonged in this population, with a mean QT corrected (QTc) of 446.5±43.2 ms. The QTc at the time of admission was similar across the world regardless of type of arrhythmia, though the maximum QTc developed during admission was highest among patients with ventricular arrhythmias.

Although the majority of patients were treated with QTc prolonging medications, including hydroxychloroquine, azithromycin, and quinolone antibiotics, there was no difference in treatment based on type of arrhythmia. Regarding antiarrhythmic therapy, amiodarone was commonly used in patients with arrhythmia, and in particular patients with tachyarrhythmias. Permanent pacemakers were uncommon and were present in about 5% of patients.

Among the 67 patients who had ventricular arrhythmias, 30 (44.8%) were specified as having monomorphic VT, and 33 (49.2%) had polymorphic VT or torsade de pointes (Table 6); the remaining 4 patients did not have the morphology of VT specified. Patients with both polymorphic VT and monomorphic VT were treated with hydroxychloroquine and azithromycin. Almost half of the patients with VT were hypoxic with a measured oxygen saturation of <90% or hypotensive with systolic blood pressure of <90 mm Hg at the time of their arrhythmia; this was similar for patients with both monomorphic and polymorphic VT. About 1 out of 3 patients were in renal failure or acidotic at the time of their arrhythmia. Patients with ventricular arrhythmias had prolonged maximum QTc; however, patients with polymorphic VT did not have significantly longer QT intervals than patients with monomorphic VT.

Table 6.

Types of Ventricular Arrhythmias in Patients With COVID-19 Worldwide

Of the 1420 patients in the cohort who died, 35 (2.4%) had VT or VF recorded as the rhythm at the time of death. The rhythm at time of death was recorded as bradycardic in 38 cases (2.7%), pulseless electrical activity in 78 patients (5.5%), and asystole in 220 patients (15.5%). The remainder (73.9%) were either not receiving rhythm monitoring at the time of death or had an unknown rhythm at the time of death. Five hundred sixty-two (39.6%) of the patients who died were mechanically ventilated during their hospitalization.

Impact of COVID-19 on Electrophysiologists Around the World

Institutions throughout the world reported drastic decreases in electrophysiology procedures including ablations and device insertions, ranging from a 20 to 95% decrease in these procedures. In Milan, a primarily cardiac hospital was completely devoted to caring for patients with COVID-19, and no electrophysiology procedures were performed. Consistent with the guidelines released by Heart Rhythm Society/American Heart Association/American College of Cardiology, most institutions reported a suspension of elective procedures and only performed procedures that were emergent or immediately life threatening which included pacemaker for complete heart block or asystole, ablations for VT storm, or lead extraction for endocarditis. We have summarized the changes in electrophysiology procedural volume and use of personal protective equipment at each center in Table II in the Data Supplement.

Discussion

In a large cohort of patients hospitalized with COVID-19 across the globe, arrhythmias were common and associated with high morbidity and mortality. Atrial arrhythmias were most common, occurring in 80% of patients with arrhythmia. Patients with cardiac arrhythmias had a high burden of medical comorbidities, although most did not have a known prior history of arrhythmia. Patients were often critically ill and had considerable mortality, with only half of those surviving to hospital discharge.

Our finding that patients with COVID-19 infection and cardiac arrhythmia had a high burden of cardiovascular comorbidities is consistent with prior reports documenting high rates of these conditions, in particular among critically ill patients.[5,12,13] A global survey of electrophysiology professionals conducted by the Heart Rhythm Society in March and April similarly found that atrial arrhythmias were the most common cardiac arrhythmia noted in patients with COVID-19.[7]

We noted significant regional differences in the incidence of these comorbidities, and, in particular, metabolic syndrome. Patients in Asia had a lower burden of these diseases, and also had a unique distribution of arrhythmia, with a lower incidence of tachyarrhythmias compared with other regions, and a more significant burden of bradyarrhythmia. This is consistent with prior analyses which have noted a lower prevalence of AF in the Asia Pacific region.[14] The reason for this regional difference is unclear, though plausibly could be related to a decreased prevalence of metabolic syndrome and obesity observed in this cohort. Both inflammation and oxidative stress have been implicated in the pathogenesis of AF and metabolic syndrome, and patients with metabolic syndrome have been shown to have a higher risk of AF.[12,13,15] Additionally, a proinflammatory state has been repeatedly implicated in the pathogenesis of severe COVID-19 infection, with several treatments proposed to mitigate the inflammatory cascade, including hydroxychloroquine, corticosteroids, and IL-6 receptor inhibition.[16-18]

COVID-19 treatment protocols varied significantly across the globe, likely due to differences in mechanisms of care, burden of illness in the community, and timing of the peak infection rate. For example, hydroxychloroquine was not standard therapy for COVID-19 in the early months of infection. However, several retrospective studies suggesting a beneficial effect of hydroxychloroquine and chloroquine were published in March and April, concurrent to the peak infection in much of Europe and the northeast United States.[19-21] The majority of patients in these regions were treated with hydroxychloroquine in our series. However, as ongoing investigations have not confirmed benefit of this treatment regimen, patients in South America may have been less likely to receive this medication given the later onset of peak infection in this region of the world. While antivirals such as oseltamivir and remdesivir were used in almost all patients in Asia and the majority of patients in Europe, patients in North and South America were infrequently treated with these medications. Although dexamethasone has recently shown some promise in treating patients hospitalized with COVID-19 infection, data regarding frequency of steroid use in this cohort was unavailable and cannot be commented on.[17]

The majority of patients with COVID-19 infection and cardiac arrhythmia were treated with systemic anticoagulation. The primary indication for anticoagulation is not known; many patients may have been anticoagulated to prevent thromboembolic events related to atrial arrhythmias. Others may have had a prior history of thromboembolism or developed acute complications in the setting of critical illness, especially given concern for a disease-related hypercoagulable state.[22,23] Indeed, during the pandemic there was interest in the utilization of therapeutic dose anticoagulation in patients critically ill with COVID-19 infection, which may account for some of the use of therapeutic anticoagulation in this population.[24,25]

At the start of the pandemic, there was substantial concern regarding the risk of ventricular arrhythmia and, in particular, torsade de pointes, given the frequent use of both hydroxychloroquine and azithromycin, both of which are known to prolong the QTc. Indeed, 40% of patients who developed polymorphic VT were treated with hydroxychloroquine and many had severe prolongation of their QTc, with a max QTc of 514.8±70.5 ms (Bazett correction). Of note, patients who developed monomorphic VT also tended to have prolonged QTc about 25% of patients were treated with hydroxychloroquine, and in both cohorts, patients commonly had reduced LV function at baseline. Taken together, this data suggest that although treatment of hydroxychloroquine may have contributed to the incidence of polymorphic VT in patients with severe COVID-19 infection, the majority of these patients were likely at high risk for this arrhythmia due to preexisting cardiac disease and critical illness. Patients with ventricular arrhythmias tended to be critically ill with a high prevalence of mechanical ventilation, renal failure, hypoxia, and severe hypotension at the time that the arrhythmia occurred.

Among those patients who died, ventricular arrhythmias were rarely noted to be the rhythm at the time of death. Only 2.4% had VT or VF recorded at the time of death, while 23.7% had nonshockable rhythms including bradycardia, pulseless electrical activity, or asystole noted at the time of death. This is suggestive of severe hypoxemia and critical illness being the primary driver of cardiac arrest in this population, rather than myocardial injury or electrical derangements leading to primary cardiac arrest.

The causes of arrhythmias in patients with COVID-19 have not been fully elucidated and are likely multifactorial. The majority of patients in our cohort with arrhythmias had comorbid conditions, including congestive heart failure and coronary artery disease, and likely were predisposed to the development of cardiac arrhythmias. The possibility of direct and indirect myocardial injury in the setting of SARS-CoV-2 infection has also been extensively discussed and may play a role in the cause of cardiac arrhythmia in this population.[12,23] Myocardial injury, as assessed by cardiac biomarkers, has been reported in up to 25% of patients with severe COVID-19 infection, and both acute myocardial injury and dysfunction associated with cardiogenic shock and myocardial injury that develops as illness severity intensifies have been described. Stress-induced myopathy, hypoxia-mediated cardiac apoptosis, proarrhythmic inflammatory state and cytokine storm, and myocardial injury due to viral invasion via angiotensin-converting enzyme 2 have all been implicated in the pathogenesis of COVID-19 associated myocardial injury.[26-29]

However, our ability to infer the importance of SARS-CoV-2 related myocardial injury on the development of cardiac arrhythmia is limited by incomplete assessment of histopathologic analysis of viral-induced injury on the cardiac conduction system as well as limited understanding of how focused or diffuse inflammation due to COVID-19 infection may affect the conduction system. Whether or not the incidence of cardiac arrhythmias is higher in patients with COVID-19 infection as compared with patients with other viral illnesses or other critically ill populations remains unclear.

Institutions that participated in this study reported significant reductions in electrophysiology procedures consistent with the guidelines released by Heart Rhythm Society/American Heart Association/American College of Cardiology. Several collaborators have published detailed reports of the effect of the COVID-19 pandemic on procedural volume. In the Lombardy and Veneto regions of Italy, rates of urgent pacemaker implantation decreased by 30% during the lockdown.[30] Ancona, Italy, reported a decrease in electrophysiology procedural volume to a median 12 procedures per week compared with a median of 24 per week in the preceding 6 month period, primarily attributed to the postponement of nonurgent interventions.[31] Consistent with our findings, a survey of 84 centers by the Italian association of Arrhythmology and Cardiac Pacing found that >90% of the centers saw a significant decrease in elective device implantation, 77% saw a significant decrease in elective ablations, 70% saw a significant decrease in emergent device implantations, and 55% saw a significant decrease in emergent ablations compared with the same period in 2019.[32]

This is the largest worldwide survey of cardiac arrhythmia occurring in patients hospitalized with COVID-19 infection. We acknowledge that our ability to make conclusions or direct comparisons is limited due to its retrospective nature, the heterogeneous data collection, and variation in the available data at each site. Additionally, some regional differences in treatment may be due to changes in our understanding of the disease process and treatment outcomes over time. This cohort was limited to cases that were submitted by each institution and often were not fully representative of all cases in that particular hospital, city, state, country, or continent. The majority of cases of arrhythmia (60%) were submitted by institutions in North America, and 40% were submitted by institutions in New York City which primarily saw patients early in the pandemic. This may skew the results and limit the generalizability of the results. Because data were collected by cardiac electrophysiologists, data in the full cohort is likely skewed towards patients with cardiac arrhythmia. Additionally, resources during the pandemic were scarce, and not all patients received 12-lead electrocardiograms or routine cardiac rhythm monitoring as part of their care, which limited the ability to identify the true burden of arrhythmia in this population. Similarly, other evaluations such as echocardiography or computed tomography were performed sparingly in this population, even in patients who were critically ill, limiting the ability to make inferences regarding the causes of the arrhythmia. Due to these limitations, we did not conduct comparative statistical analysis to avoid making comparisons among heterogeneous groups.

Conclusions

COVID-19 has had a profound effect on the lives of millions of people across the globe. Many patients hospitalized with COVID-19 infection developed cardiac arrhythmia, which was associated with high morbidity and mortality and highlights the need for electrophysiologists to be involved in COVID-19 care. Regional differences in the type of arrhythmias affecting these patients may provide important insights into the pathophysiology of both SARS-CoV-2 infection and of the arrhythmia itself. Future investigations regarding the mechanisms of the cardiac complications of SARS-CoV-2 infection may aid in our understanding of arrhythmia in this population.

Acknowledgments

The Columbia University group thank Lauren Privitera, MS, MPH, Director of Cardiology Research for her help in setting up our RedCAP database. Figure was created with BioRender.com. We would like to acknowledge all the worldwide health care providers including doctors, nurses, medical assistants, physician assistants and their selflessness in taking care of patients with coronavirus disease (COVID).

Sources of Funding

Dr Rodriguez is funded by a career development award from the National Heart, Lung, and Blood Institute (K01 HL 14460) and the American Heart Association/Robert Wood Johnson Harold Amos Medical Faculty Development Program. Dr Wan was supported by National Institutes of Health (NIH) R01 HL152236, the Esther Aboodi Endowed Professorship at Columbia University, and Wu Family Research Fund.

Disclosures

Dr Biviano has served on medical advisory boards for Boston-Scientific, Biosense Webster, and Abbott, not relevant to this work. The other authors report no conflicts.