Anticoagulation Initiation in Critically Ill Patients With New-Onset Atrial Fibrillation and Sepsis

Introduction: Patients in the medical intensive care unit (MICU) with sepsis have a higher risk of developing new onset atrial fibrillation (NOAF). Several risk factors increase the risk of stroke in this population including activation of the coagulation cascade in sepsis. Simultaneously, critically ill patients often harbor increased risks for bleeding due to comorbidities in the setting of organ dysfunction. Existing risk stratification tools to evaluate the safety and efficacy of anticoagulation are not validated in critically ill patients.


Methods: This was a descriptive, retrospective cohort study of MICU patients admitted to a 1,000-bed tertiary community teaching hospital between 1/1/2019 and 8/1/2024. Patients ≥ 18 years of age with a diagnosis of NOAF and sepsis during MICU admission were included. Patients were excluded for a history of anticoagulation use or cardiac surgery within 3 months. The primary aim was to compare the percentage of patients with NOAF and sepsis who were initiated on anticoagulation versus those not initiated on anticoagulation in the MICU. Secondary outcomes included the incidence of thromboembolic events, major bleeding, and death in the patients initiated on anticoagulation, and the clinical characteristics associated with anticoagulation.


Results: A total of 320 patients were screened, and 44 patients met study inclusion criteria. The mean age of the patients was 69.5 (SD 14.9) years of age and 59% of patients were male. Of the 44 included patients, 29 were initiated on anticoagulation and 4 were continued on anticoagulation at discharge. Additionally, 62% of patients initiated on anticoagulation died in the MICU. The median CHA2DS2-VASc score was 5 (0-7) and the median HAS-BLED score was 4 (1-6). While no patients experienced pulmonary embolism, 2 patients had a deep vein thrombosis, and 3 patients experienced an embolic stroke. None of the patients experienced a major bleeding event.


Conclusions: Due to the elevated risk of experiencing a stroke in this patient population coupled with the relatively low risk of bleeding illustrated in this study, it may be reasonable to initiate anticoagulation in patients with sepsis that develop NOAF. However, further research is warranted to improve understanding of the patients that would benefit most from anticoagulation.