An early evaluation of bleeding-related hospital readmissions among hospitalized patients with nonvalvular atrial fibrillation treated with direct oral anticoagulants

Steve Deitelzweig, Amanda Bruno, Jeffrey Trocio, Natalie Tate, Kiran Gupta, Jay Lin, Melissa Lingohr-Smith

Clinical trials have demonstrated direct oral anticoagulants (DOACs) are efficacious for reducing stroke risk among patients with nonvalvular atrial fibrillation (NVAF) with differences in the reduction of bleeding risks vs. warfarin. The objective of this study was to assess bleeding-related hospital readmissions among hospitalized NVAF patients treated with dabigatran, rivaroxaban, and apixaban, in the US. Patients (≥18 years) with a discharge diagnosis of NVAF who received apixaban, dabigatran, or rivaroxaban during hospitalization were identified from the Premier Hospital database (1/1/2012-3/31/2014) and the Cerner Health Facts hospital database (1/1/2012-8/31/2014). Patients identified from each database were analyzed separately and grouped into 3 cohorts depending on which DOAC was received. Patient characteristics, hospital resource use and costs, and frequency of readmissions within 1 month were evaluated. Among study populations identified from the Premier database (N=74,730) and the Cerner database (N=14,201), patients who received apixaban were older, had greater comorbidity, and had higher stroke and bleeding risks. After controlling for patient characteristics, including comorbidity and stroke and bleeding risks, compared with patients who received apixaban during their index hospitalizations, the odds of bleeding-related hospital readmissions were significantly greater by 1.4-fold (p<0.01) for patients who received rivaroxaban and 1.2-fold (p=0.16) numerically greater for patients who received dabigatran among patients identified from the Premier Hospital database. Among patients in the Cerner Health Facts hospital database, bleeding-related hospital readmissions were significantly greater by 1.6-fold (p=0.04) for patients who received rivaroxaban and 1.3-fold (p=0.30) numerically greater for patients who received dabigatran compared to patients who received apixaban. No causal relationship between treatment and outcomes can be concluded. NVAF patients using different DOACs had different characteristics, including stroke and bleeding risks. Use of rivaroxaban, compared to apixaban was associated with significantly greater risk of bleeding-related readmissions across two database claims analyses.