Patterns of Prophylactic Anticoagulation Among Patients Hospitalized for COVID-19: An Analysis of the American Heart...

Limited knowledge exists about prophylactic anticoagulation patterns in patients hospitalized for COVID-19.

We conducted a retrospective cohort study using American Heart Association COVID-19 Cardiovascular Disease Registry data from May 2020 to March 2022. We included patients without preexisting indications for or contraindications to anticoagulation, excluding those with missing anticoagulation data. Patients were categorized by the highest anticoagulation dose received. Multilevel logistic regression was used to assess the relationship between anticoagulation use/dose, patient demographics, clinical presentation, in-hospital course, institutional characteristics, and admission date, accounting for hospital clustering. Among 26 775 patients, 4157 (16%) received no anticoagulation, 15 617 (58%) low-dose, 3071 (11%) intermediate-dose, and 3930 (15%) full-dose anticoagulation. Significant hospital-level variability occurred for any anticoagulation use (range, 0%-98%; P<0.0001) and by dose (full anticoagulation range, 0%-85%; P<0.0001). Controlling for hospital variability, older age, male sex, non-White race, higher body mass index, higher platelets, corticosteroid use, and intensive care unit admission were positively associated with any anticoagulation use. Older age, male sex, higher body mass index, higher platelets, corticosteroid use, intensive care unit admission, mechanical ventilation, and admission before October 2020 were associated with higher anticoagulation dose (full versus low dose). Rates of no anticoagulation significantly increased in both intensive care unit and non-intensive care unit strata over time (P trend=0.01 and <0.0001, respectively).

In this large real-world analysis, nearly 1 in 6 patients hospitalized for COVID-19 received no prophylactic anticoagulation. Patient and disease characteristics associated with thrombotic risk and COVID-19 severity correlated with anticoagulation strategy. Importantly, substantial institutional differences emerged, highlighting gaps between clinical practice and guideline recommendations.