Impacts of hemoglobin levels and platelet counts on clinical outcomes in patients with atrial fibrillation: oral anticoagulant use for patients with anemia and/or thrombocytopenia

Abstract Background Hemoglobin (Hgb) levels and platelet (PLT) counts have beeen associated with adverse clinical outcomes in some patients with cardiovascular conditions. We aimed to assess the risks of clinical events among patients with atrial fibrillation (AF) in relation to Hgb levels and PLT counts. Second, we investigated clinical outcomes on warfarin or non-vitamin K antagonist oral anticoagulants (NOACs) compared to no oral anticoagulant use (non-OAC), in different Hgb and PLT strata. Methods We used medical data from a multi-center healthcare system in Taiwan which included 37,074 AF patients. Patients were categorized into 3 groups based on their Hgb levels and PLT counts: Group 1 (Hgb>10g/dL and PLT>100K/uL; n=29,147), Group 2 (Hgb<10g/dL or PLT<100K/uL; n=7,078) and Group 3 (Hgb<10g/dL and PLT<100K/uL; n=849). Patients in each category were further stratified as 3 groups according to their stroke prevention strategies; that is, non-OAC, warfarin or NOACs. Results A higher Hgb level and/or PLT count was associated with a higher risk of ischemic stroke/systemic embolism (IS/SE), but lower risks of intracranial hemorrhage (ICH) and major bleeding. The composite risks of IS/SE, ICH and major bleeding were higher in Group 3 or Group 2, compared to Group 1 (6.75%/yr versus 6.41%/yr versus 4.13%/yr). Compared to non-OACs, the use of warfarin was not associated with a lower composite risk of IS/SE, ICH and major bleeding in the 3 groups. NOACs were associated with a lower composite risk in Group 1 (aHR 0.68, 95% CI 0.60–0.76) and Group 2 (aHR 0.73, 95% CI 0.53–0.99), but was non-significant in Group 3 (aHR 0.73, 95% CI 0.17–3.07) (Figure). The net clinical benefits were consistently positive in different weight models, in favor of NOAC use, in Group 2 and its subgroups with either anemia or thrombocytopenia. Conclusions AF patients with anemia and/or thrombocytopenia were a high-risk population. Compared to no OAC use, NOACs were associated with significantly better clinical outcomes for patients with advanced anemia (Hgb<10g/dL) or thrombocytopenia (PLT<100K/uL), but not for those with both conditions. Harms or benefits of NOACs should be carefully evaluated and balanced in this population. Funding Acknowledgement Type of funding sources: None.