Arrhythmia is a potential cardiovascular complication of Coronavirus Disease 2019 (COVID-19) [1]. In one case series of patients hospitalized with COVID-19, 16.7% developed unspecified arrhythmia [2], while another case series indicated sustained ventricular tachycardia or ventricular fibrillation among 5.9% of patients hospitalized with COVID-19 [3]. However, incidence rates of arrhythmia and mortality rates after incident arrhythmia in COVID-19 patients have not been systematically established.

We searched for relevant studies cited in PubMed or Embase up to September 15, 2020, using the terms “COVID-19”, “arrhythmia”, “incidence”, “mortality,” and “prognosis” with suitable MeSH terms. All studies were selected and reviewed by two reviewers (SCL and SCS). The final list of included studies and data extractions were derived through extensive discussion with agreement from both authors. Statistical analyses were performed using MedCalc (Windows) version 15.0 (MedCalc Software, Ostend, Belgium). Outcomes were reported as proportions with 95% confidence interval (CI), based on the random effects model. The heterogeneity among studies was detected by the Cochran Q test with p value and the I2 statistic.

Of 645 potential studies screened, we excluded 143 duplicate studies, 66 irrelevant studies, 12 conference abstracts, 241 other types of publications (e.g., pre-prints, protocols, opinions, recommendations, editorials, commentaries, retractions and reviews), 114 studies without incidence or mortality data, and 13 non-English studies. We included 56 studies from 11 countries comprising 17,435 patients with COVID-19. Study characteristics for included articles are listed in Table 1. Notably, most studies only included hospitalized patients with COVID-19 (96.4%). The overall incidence of arrhythmia in COVID-19 patients was 16.8% (95% CI: 12.8–21.2%; I2: 98.0%, p < 0.001) (Fig. 1a). The incidence of different types of arrhythmia in patients with COVID-19 was as follows: 12.0% (22 studies, 95% CI: 8.6–15.9%) for non-classified arrhythmia, 8.2% (14 studies, 95% CI: 5.5–11.3%) for atrial fibrillation/atrial flutter/atrial tachycardia, 10.8% (26 studies, 95% CI: 6.6–15.9%) for conduction disorders, 8.6% (5 studies, 95% CI: 4.5–13.9%) for premature contraction and 3.3% (16 studies, 95% CI: 1.9–4.9%) for ventricular fibrillation/ventricular tachycardia. We found the mortality was 20.3% (95% CI: 12.9–29.0%; I2: 72.8%, p < 0.001) in COVID-19 patients who developed arrhythmia (Fig. 1b).

Table 1 Study characteristics
Full size table
Fig. 1
figure 1

Forest plot of a arrhythmia incidence in COVID-19 infections and b mortality in COVID-19 patients with incident arrhythmia from included studies

Full size image

Compared to the incident arrhythmia in patients with community-acquired pneumonia (4.7%, 95% CI: 2.4–8.9) [4], the present study indicates higher incidence of arrhythmia in COVID-19 patients (16.8%) with 2 out of 10 patients dying after developing arrhythmia. The possible mechanisms of arrhythmia may involve cardiac damage from metabolic disarray, hypoxia, neuro-hormonal or inflammatory stress and infection-related myocarditis in the setting of COVID-19 [5]. Notably, higher incidence rates of conduction disorders and premature contraction were found in COVID-19 patients, compared to other types of arrhythmia, in the present study. Our findings increase clinical awareness of arrhythmia in patients hospitalized with COVID-19 for the benefit of first-line healthcare providers.

The major limitation of our study was the inclusion of studies largely from observational data with the potential risk of selection bias. For example, nearly all included studies analyzed data from inpatient settings rather than from the community, likely resulting in overestimation of the true incidence and mortality of arrhythmia among COVID-19 infections. In addition, heterogeneity within and between countries may have caused differences in the estimated incidence and clinical impacts of arrhythmia. Finally, due to the involvement of multiple factors, mortality in COVID-19 patients who developed arrhythmia cannot be entirely attributed to arrhythmia alone. However, the strength of the present study is to summarize the current evidence regarding arrhythmia and COVID-19 infection from various populations worldwide. Since COVID-19 infection probably poses increased risk of arrhythmia, significantly affecting mortality, physicians should consider arrhythmia monitoring with early management in addition to supportive care and respiratory support when treating COVID-19 patients.