The increasing use of direct oral anticoagulants (DOACs) has resulted in more anticoagulated patients presenting with acute upper gastrointestinal bleeding (UGIB). This study aimed to compare clinical presentation, endoscopic findings, interventions, and outcomes between DOAC users and non-anticoagulated patients presenting with acute UGIB.

We conducted a retrospective review of all UGIB cases undergoing endoscopy at Sheffield Teaching Hospitals between January and December 2023. Patients were categorized into DOAC users and non-users. Data collected included demographics, presenting symptoms, endoscopic diagnoses, therapeutic interventions, length of hospital stay, and short-term mortality. Comparative analyses were performed to identify differences between the two groups.

A total of 649 patients were included, of whom 98 (15.1%) were taking DOACs (Apixaban n=52, Edoxaban n=27, Rivaroxaban n=19) and 551 (84.9%) were not. DOAC users were significantly older than non-DOAC patients, with a mean age of 75.0 years (95% CI 72.4–77.6) compared with 55.9 years (95% CI 54.3–57.5; p < 0.001). The principal indications for DOAC therapy were atrial fibrillation (n=69, 70.4%) and thrombotic disease (n=29, 29.6%). Melena at presentation was reported in 43/98 (43.9%) of DOAC users compared with 198/551 (35.9%) non-DOAC patients. Common endoscopic findings in DOAC versus non-DOAC patients included oesophagitis (17/98, 17.3% vs 84/551, 15.2%), gastric ulcers (11/98, 11.2% vs 40/551, 7.3%), duodenal ulcers (5/98, 5.1% vs 43/551, 7.8%), and varices (4/98, 4.1% vs 66/551, 12.0%); however, these differences did not reach statistical significance (all p > 0.05). Therapeutic endoscopic procedures were less commonly performed in DOAC users compared with non-DOAC patients (8/98, 8.2% vs 61/551, 11.1%). (p=0.48). Interventions in the DOAC group included injection therapy (n=3), clips (n=2), EMR for a bleeding polyp (n=1), Bentonite haemostatic agent (n=1), and other modalities (n=3), whereas the non-DOAC group underwent APC (n=2), band ligation (n=19), injection therapy (n=19), clips (n=22), EMR (n=3), heater probe (n=5), Bentonite haemostatic agent (n=3), and other treatments (n=7). Length of hospital stay was significantly longer among DOAC users (6.0 vs 3.6 days; p = 0.009). One death within 10 days of admission occurred in the DOAC group.

Patients presenting with UGIB while on DOAC therapy were older, underwent fewer therapeutic endoscopic interventions, and experienced longer hospital admissions compared with non-anticoagulated patients. Although endoscopic findings were similar, the extended hospital stay highlights the added complexity of managing anticoagulated patients. These findings emphasize the need for tailored management pathways and resource planning to optimize outcomes in this growing patient population.