Predictors of arrhythmias in the population hospitalized for SARS-CoV-2

Abstract

Background

Studies exploring predictors of arrhythmias in the population primarily hospitalized for SARS-CoV-2 (COVID-19) are scarce. Understanding this is crucial for risk stratification and appropriate management.

Methods

Using the 2020 National Inpatient Sample (NIS) database, we identified primary admissions for COVID-19. A ‘greedy neighbor’ 1:1 propensity-score matching (PSM) accounted for baseline differences. Then, multivariable logistic regression models were employed to account for confounders and estimate the probability of arrhythmia.

Results

There were a total of 1,058,815 admissions for COVID-19 (mean age 64.3 years ±16.8), 47.2% female, 52.5% (107698) White, 18.5% (37973) Blacks, and 20.7% (42,447) Hispanics. Atrial fibrillation was the most prevalent arrhythmia, 15.1% (31,942). After PSM, 166,405 arrhythmia hospitalizations were matched to 166,405 hospitalizations without arrhythmia. Sick sinus syndrome 4.9 (4.4-5.5), dyslipidemia 1.2 (1.2–1.3), cardiac arrest 1.3 (1.1-1.4), invasive mechanical ventilation 1.9 (1.8-2.0) and obesity 1.3 (1.2-1.4), (p<0.0001, all) were all independent predictors of arrhythmias.

Conclusions

Our analysis revealed a notable proportion of hospitalized COVID-19 patients with arrhythmias. Dyslipidemia, obesity, sick sinus syndrome, invasive mechanical ventilation, and cardiac arrest were independent predictors of arrhythmias.