Preadmission Oral Anticoagulation for Atrial Fibrillation/Flutter and Death or Thrombotic Events During COVID-19 Admission

Abstract

Atrial fibrillation/flutter (AF) and COVID-19 are associated with an elevated risk of arterial and venous thrombosis. Whether preadmission oral anticoagulation (OAC) for AF reduces the incidence of in-hospital death or thrombotic events among patients with COVID-19 is unknown. We identified 630 patients with pre-existing AF and a hospitalization diagnosis of COVID-19 and stratified them according to preadmission OAC use. Multivariable logistic regression was employed to relate preadmission OAC to composite in-hospital mortality or thrombotic events. Unadjusted composite in-hospital mortality or thrombotic complications occurred less often in those on than not on preadmission OAC (27.1% vs 46.8%, p <0.001). After adjustment, the incidence of composite in-hospital all-cause mortality or thrombotic complications remained lower with preadmission OAC (odds ratio 0.37, confidence interval 0.25 to 0.53, p <0.0001). Secondary outcomes including all-cause mortality (16.3% vs 24.9%, p = 0.007), intensive care unit admission (14.7% vs 29.0%, p <0.001), intubation (6.4% vs 18.6%, p <0.001), and noninvasive ventilation (18.6% vs 27.5%, p = 0.007) occurred less frequently, and length of stay was shorter (6 vs 7 days, p <0.001) in patients on than those not on preadmission OAC. A higher CHA₂DS₂-VASc score was associated with an increased risk of thrombotic events. In conclusion, among patients with baseline AF who were hospitalized with COVID-19, those on preadmission OAC had lower rates of death, arterial and venous thrombotic events, and less severe COVID-19.

Figure 1

Study Flowchart. 3,584 patients were admitted with COVID-19 between 3/1/20–12/31/20. After applying exclusion criteria, 630 of these had atrial fibrillation or flutter (AF), of which 361 (57%) were single-agent OAC users, and 269 (43%) were non-OAC users.

Figure 2

Multivariable Model of All-Cause Mortality and Thrombotic Outcomes in the Overall Cohort. Shown is a multivariable model relating pre-admission use of oral anticoagulation to composite in-hospital all-cause mortality and thrombotic complications.Abbreviations: RAAS = renin angiotensin aldosterone system. “Admission date +30” was defined as the index date of admission plus 30 days to evaluate the temporal effect of admission on outcomes.