Network meta-analysis compares DOACs and VKAs for intracranial hemorrhage risk

Photo by National Cancer Institute / Unsplash

Frontiers in Medicine Published May 19, 2026 Medically reviewed May 19, 2026 DOI ↗ By Dr. Amelia Tan, PhD · Internal Medicine & Chronic Disease

Key Takeaway

Consider that all DOACs lower ICH risk vs VKAs; dabigatran ranked safest but SUCRA is not definitive.

This network meta-analysis included 127,267 patients with venous thromboembolism or atrial fibrillation and compared direct oral anticoagulants (DOACs) with Vitamin K Antagonists (VKAs) for the risk of intracranial hemorrhage (ICH). The analysis included apixaban, dabigatran, edoxaban, rivaroxaban, and VKAs.

Key findings showed that VKAs had higher risk of ICH than apixaban (OR 2.40, 95% CI 1.62–3.56). Rivaroxaban and VKAs had higher risk than dabigatran (rivaroxaban vs dabigatran: OR 1.89, 95% CI 1.18–3.04; VKAs vs dabigatran: OR 2.83, 95% CI 2.00–3.99). Similarly, rivaroxaban and VKAs had higher risk than edoxaban (rivaroxaban vs edoxaban: OR 1.60, 95% CI 1.04–2.47; VKAs vs edoxaban: OR 2.39, 95% CI 1.81–3.17). VKAs also had higher risk than rivaroxaban (OR 1.50, 95% CI 1.08–2.07).

Safety ranking by SUCRA placed dabigatran highest (87.6), followed by apixaban (68.6), edoxaban (67.5), rivaroxaban (26.1), and VKAs (0.2). However, SUCRA is a ranking probability and not a direct comparison; superiority of one DOAC over another should not be claimed based on SUCRA alone.

Limitations were not reported in the source. The results are from a network meta-analysis of RCTs, providing association only. For practice, all DOACs had a lower risk of ICH than VKAs, and dabigatran may be the safest choice regarding ICH risk, but cautious interpretation of rankings is warranted.

Study Details

Study typeMeta analysis

EvidenceLevel 1

PublishedMay 2026

View Original Abstract ↓

ObjectiveWe updated a network meta-analysis of randomized controlled trials to compare the risk of intracranial hemorrhage (ICH) between direct oral anticoagulants (DOACs) and Vitamin K Antagonists (VKAs) in detail across Venous thromboembolism and atrial fibrillation.MethodsPubMed, EMBASE, Web of Science, and the Cochrane Library databases were searched up to January 5, 2026. The incidence of ICH was investigated. Using frequentist network meta-analysis, interventions that were not compared directly could be compared indirectly by the 95% confidence interval (CI), making the search results more intuitive. Based on surface under the cumulative ranking curves (SUCRA), the relative ranking probability of each group was generated.ResultsTwenty randomised controlled trials (127,267 patients) were included. Compared with apixaban, VKAs (OR: 2.40, 95% CI: 1.62–3.56) had a higher risk of bleeding, and the difference was significant. Compared with dabigatran, rivaroxaban (OR: 1.89, 95% CI: 1.18–3.04) and VKAs (OR: 2.83, 95% CI: 2.00–3.99) had a higher risk of bleeding, and the difference was significant. Compared with edoxaban, rivaroxaban (OR: 1.60, 95% CI: 1.04–2.47) and VKAs (OR: 2.39; 95% CI: 1.81–3.17) had a higher risk of bleeding, and the difference was significant. Compared with rivaroxaban, VKAs (OR: 1.50; 95% CI: 1.08–2.07) had a higher risk of bleeding, and the difference was significant. In the ranking of the cumulative probability of ICH, dabigatran (SUCRA 87.6) had the highest safety, followed by apixaban (SUCRA 68.6), edoxaban (SUCRA 67.5), rivaroxaban (SUCRA 26.1), and VKAs (SUCRA 0.2).ConclusionsAll DOACs had a lower risk of ICH than VKAs. Dabigatran may be the safest choice among any anticoagulant regarding risk of ICH.Systematic Review Registrationhttps://www.crd.york.ac.uk/PROSPERO/view/CRD420261336668, identifier CRD420261336668.