Safety of Direct Oral Anticoagulants for Gastrointestinal Hemorrhage in Patients With Nonvalvular Atrial Fibrillation: A Systematic Review and Meta-analysis of Real-world Studies.

Abstract

Since the introduction of Direct Oral Anticoagulants (DOACs), "real-world" studies have investigated their safety profile on gastrointestinal hemorrhage (GIH) when used by patients with Non-Valvular Atrial Fibrillation. We performed a systematic review and meta-analysis to compile and summarize this data after Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. Medline and Embase were systematically searched until April 2021. Observational studies that met predefined inclusion criteria were included and hazard ratios (HRs) with 95% CI were extracted. Subgroup analyses based on DOAC doses, history of chronic kidney disease, stroke, prior exposure to VKA (vitamin K antagonist), age, gender, geographic location of population samples, as well as Leave-One-Out and Low/Moderate Risk of Bias sensitivity analyses were performed. A random effects model was used. A total of 46 studies were included. Apixaban was associated with a reduced risk of GIH compared with Dabigatran (HR: 0.67, 95% CI, 0.56 to 0.81, I2 : 53.28%), Rivaroxaban (HR: 0.56, 95% CI, 0.44 to 0.70, I2 : 79.17%), and VKA (HR: 0.68, 95% CI, 0.60 to 0.78, I2 : 71.93%). Rivaroxaban was associated with increased GIH risk compared with Dabigatran (HR: 1.19, 95% CI, 1.02 to 1.40, I2 : 72.96%) and VKA (HR: 1.16, 95% CI, 1.05 to 1.27, I2 : 81.95%). Dabigatran was associated with similar GIH risk compared with VKA (HR: 1.11, 95% CI, 0.98 to 1.26, I2 : 87.28%). Our study shows that Apixaban was associated with a reduction in GIH risk compared with Dabigatran, Rivaroxaban and VKA, whereas Rivaroxaban was associated with an increase in GIH risk compared with both Dabigatran and VKA.