Identifying the bleeding source in obscure gastrointestinal bleeding is often challenging, particularly when endoscopic, radiologic, and surgical explorations yield discordant results. We present a case of recurrent bleeding of uncertain etiology, despite extensive investigation and therapeutic interventions.

A 73-year-old male with atrial fibrillation (previously anticoagulated), Child A cirrhosis, and recurrent C. Difficile colitis presented with repeated episodes of hematochezia and melena and moderate microcytic anemia.

Initial EGD revealed only small esophageal varices without signs of recent bleeding. Colonoscopy identified a cecal polyp which was resected, and no evidence of active bleeding. During an episode of hematochezia, we performed emergency videocapsule endoscopy, which showed the presence of blood in the terminal ileum but subsequent angio-CT was negative.

Due to recurrent episodes of clinically significant bleeding, we opted for intraoperative enteroscopy, which identified several small-bowel angiodysplasias, subsequently treated with APC and clipping. However, these lesions were small, of limited bleeding potential, and did not fully correlate with the severity and recurrence of overt bleeding. Recurrent melena persisted despite therapy, requiring multiple transfusions.

Bleeding gradually decreased after supportive management and repeated procedures, and the patient was discharged clinically stable. However, the definitive source of bleeding remained unclear.

Although multiple angiodysplasias were identified, their significance as the primary bleeding source is uncertain. A small-bowel Dieulafoy-type lesion - intermittent and difficult to capture - remains a strong alternative explanation. This case highlights the inherent limitations of current diagnostic modalities in obscure GI bleeding and the need for repeated, multimodal evaluation.