Comparative outcomes of direct oral anticoagulants versus warfarin in patients with atrial fibrillation and liver cirrhosis: A propensity score matched TriNetX analysis

Anticoagulation in patients with atrial fibrillation (AF) and liver cirrhosis is challenging due to competing risks of thrombosis and bleeding. While direct oral anticoagulants (DOACs) have emerged as alternatives to warfarin, their comparative safety and effectiveness in cirrhosis remain incompletely defined. We conducted a retrospective cohort study using the TriNetX Global Collaborative Network. Adult patients (≥ 18 years) with AF and cirrhosis receiving DOACs or warfarin were identified. Propensity score matching (1:1) was performed to balance baseline characteristics. The primary outcomes were hepatic decompensation and ischemic stroke within 1 year. Secondary outcomes included all-cause mortality, major bleeding, intracranial hemorrhage, gastrointestinal bleeding, and all-cause hospitalization. Survival (Kaplan-Meier) analysis and hazard ratios were estimated. After propensity score matching, 2,522 matched pairs (5,044 patients) were included. Over 1-year follow-up, hepatic decompensation occurred in 8.8% of DOAC patients and 11.1% of warfarin patients, with a significantly lower risk observed in the DOAC group (HR 0.79, 95% CI 0.67-0.95; p = 0.010). Ischemic stroke occurred in 2.9% of DOAC patients and 2.4% of warfarin patients, with no statistically significant difference between groups (HR 1.22, 95% CI 0.85-1.75; p = 0.289). All-cause mortality was lower in the DOAC group (17.7% vs. 20.3%; HR 0.87, 95% CI 0.77-0.99; p = 0.035). No significant differences were observed in major bleeding (HR 0.84, 95% CI 0.70-1.02), intracranial hemorrhage (HR 0.98, 95% CI 0.59-1.62), gastrointestinal bleeding (HR 0.97, 95% CI 0.77-1.23), or all-cause hospitalization (HR 1.04, 95% CI 0.97-1.12). In patients with atrial fibrillation and cirrhosis, DOAC use was associated with lower risks of hepatic decompensation and all-cause mortality compared with warfarin, while no significant difference was observed in ischemic stroke or other safety outcomes. Given the observational design and modest absolute differences, these findings should be interpreted as associative and hypothesis-generating rather than causal or definitive.