Outcomes of Anemic Versus Non-Anemic Patients Anticoagulated For Atrial Fibrillation and/or Venous Thromboembolic Disease

Document Type

Conference Proceeding

Publication Date

9-2026

Publication Title

Research and Practice in Thrombosis and Haemostasis

Abstract

Background Anemia has been associated with bleeding outcomes for anticoagulated patients. There are several potential mechanisms for this, including hemorheology and red blood cells impacting hemostasis. Aims To evaluate long-term bleeding outcomes and healthcare utilization for patients on anticoagulation with versus without baseline anemia. Methods Data from November 2015-December 2024 were analyzed from the Michigan Anticoagulation Quality Improvement Initiative registry. Patients on long-term anticoagulation with a direct oral anticoagulant for atrial fibrillation and/or venous thromboembolism with at least 6 months of follow-up data were included; patients on dialysis or receiving chemotherapy were excluded. Patients were classified as anemic (defined as hemoglobin less than 12.0 g/dL or hematocrit < 36% for women and hemoglobin < 13.5 g/dL or hematocrit < 40% for males), non-anemic, or missing data based on complete blood count monitoring up to six months before starting anticoagulation. Patients underwent 1:1 propensity score matching based on demographics, comorbidities, medications, indication for anticoagulation, recent bleeding or surgery, prescribing doctor, and follow-up. Using baseline hemoglobin values, anemic versus non-anemic patients were compared. The primary outcome was the rate of bleeding. Secondary outcomes included rates of bleeding subtypes, and healthcare utilization. Results Of 1,986 patients without anemia, 1,311 baseline anemic patients, and 380 missing data patients, mean (± standard deviation) age was 70.3 years (±12.7), 51.2% were male sex, 81.3% white, and 71.9% were anticoagulated for atrial fibrillation. Patients were followed for an average of 33.6 months (Table 1). Unadjusted data showed baseline anemia patients to have a higher rate of bleeding (26.7 versus 19.9 versus 19.8 per 100 patient years, p< 0.0001), major bleeding (3.8 versus 2.0 versus 2.0 per 100 patient years, p< 0.0001), surgery, endoscopy, emergency department visits, hospitalization, and mortality compared to the non-anemic and missing baseline data groups, respectively. After propensity score matching patients anemic at baseline compared to those with a normal baseline hemoglobin (n=1,298 in each group), these differences were no longer statistically significant (table 2). Conclusions Unadjusted, baseline anemia was associated with adverse bleeding outcomes, increased healthcare utilization, and mortality for anticoagulated patients. However, there was no statistically significant difference in outcomes after propensity matching.

Volume

10

Issue

Suppl 2

First Page

103869

Last Page

103869

Comments

International Society on Thrombosis and Haemostasis (ISTH) Congress, July 11-15, 2026, Paris, France

DOI

10.1016/j.rpth.2026.103869

Link to Full Text

Share

COinS