RESUMO
Gastrointestinal bleeding (GIB) is the most common type of bleeding occurring in patients on oral anticoagulation. A meta-analysis of the landmark randomized controlled trials (RCTs) for patients with atrial fibrillation demonstrated that direct oral anticoagulants (DOACs) were associated with higher GIB rates compared to warfarin. However, significant heterogeneity existed between studies. While rivaroxaban, high-dose dabigatran, and high-dose edoxaban were associated with higher GIB rates than warfarin, GIB rates were similar between warfarin users and both apixaban and low-dose dabigatran users. Additionally, previous observational studies have yielded conflicting reports on whether GIB rates differ between warfarin and DOACs. Meta-analyses of observational studies demonstrated that warfarin is associated with lower rates of GIB compared to rivaroxaban, similar or lower rates compared to dabigatran, and higher rates compared to apixaban. Importantly, no RCT has compared individual DOACs directly and due to the different selection criteria of the initial RCTs, indirect comparisons between DOACs using these studies are unreliable. The best available information of comparisons between individual DOACs is therefore limited to observational studies. There is mounting evidence that suggests that rivaroxaban is associated with a higher risk of GIB compared to other DOACs. Finally, GIB induced by oral anticoagulation may have some positive aspects. Interestingly, there are studies that indicate oral anticoagulation facilitates colorectal cancer detection. Furthermore, results from RCTs and observational studies suggest that warfarin may even decrease the incidence of cancer.