BACKGROUND: Evidence on the risk of gastrointestinal (GI) bleeding associated with dabigatran etexilate (DE) is contrasting. We performed a meta-analysis of literature to address this issue. METHODS AND RESULTS: Studies on GI bleeding risk in patients receiving DE or vitamin-K antagonists (VKA) were systematically searched. Twenty-three studies (26 datasets) showed no difference in the GI bleeding risk between the 250,871 patients treated with DE and the 460,386 receiving VKA (OR: 1.052, 95% CI: 0.815, 1.359). Similar results were obtained when pooling together adjusted ORs/HRs, obtained by means of multivariate analysis (OR: 1.06, 95% CI: 0.914, 1.222). Compared with VKA, DE use was associated with a significantly lower risk of upper GI (OR: 0.742, 95% CI: 0.569, 0.968), but not of lower GI bleedings (OR: 1.208, 95% CI: 0.902, 1.619). Furthermore, no significant difference in the GI bleeding risk was found when data on DE 110 mg and DE 150 mg twice-daily were separately compared with VKA. CONCLUSIONS: No difference in GI bleeding risk was found between DE and VKA. These results were confirmed for both dosages of DE and when specifically analyzing lower GI bleeding. In contrast, the risk of upper GI bleeding was lower with DE than with VKA. KEY MESSAGES No difference in the risk of gastrointestinal (GI) bleeding can be found between dabigatran etexilate (DE) and vitamin K-antagonists (VKA). These results are confirmed for both dosages of DE. The risk of upper GI bleeding is lower with DE than with VKA.
BACKGROUND: Evidence on the risk of gastrointestinal (GI) bleeding associated with dabigatran etexilate (DE) is contrasting. We performed a meta-analysis of literature to address this issue. METHODS AND RESULTS: Studies on GI bleeding risk in patients receiving DE or vitamin-K antagonists (VKA) were systematically searched. Twenty-three studies (26 datasets) showed no difference in the GI bleeding risk between the 250,871 patients treated with DE and the 460,386 receiving VKA (OR: 1.052, 95% CI: 0.815, 1.359). Similar results were obtained when pooling together adjusted ORs/HRs, obtained by means of multivariate analysis (OR: 1.06, 95% CI: 0.914, 1.222). Compared with VKA, DE use was associated with a significantly lower risk of upper GI (OR: 0.742, 95% CI: 0.569, 0.968), but not of lower GI bleedings (OR: 1.208, 95% CI: 0.902, 1.619). Furthermore, no significant difference in the GI bleeding risk was found when data on DE 110 mg and DE 150 mg twice-daily were separately compared with VKA. CONCLUSIONS: No difference in GI bleeding risk was found between DE and VKA. These results were confirmed for both dosages of DE and when specifically analyzing lower GI bleeding. In contrast, the risk of upper GI bleeding was lower with DE than with VKA. KEY MESSAGES No difference in the risk of gastrointestinal (GI) bleeding can be found between dabigatran etexilate (DE) and vitamin K-antagonists (VKA). These results are confirmed for both dosages of DE. The risk of upper GI bleeding is lower with DE than with VKA.