No catheter angiography is needed in patients with an obscure acute gastrointestinal bleed and negative CTA

Purpose To evaluate the negative predictive power of computed tomography angiography (CTA) for the identification of obscure acute gastrointestinal (GI) bleeding (GI bleeding not visualized/treated by endoscopy) on subsequent mesenteric angiography (MA) with the intention to treat. Materials and methods A retrospective chart review of patients was performed who underwent mesenteric angiography for the evaluation/treatment of acute GI bleeding between November 2012 and July 2016. Patients with negative CTA examinations that proceeded to MA were identified. Negative predictive value (NPV) was calculated. Results 20 patients (14 male, 6 female; average age: 73.1 ± 12.8 years) underwent 20 negative CTA examinations for the evaluation and treatment of GI bleeding followed by mesenteric angiography. Eighteen of 20 patients had negative subsequent MA (negative predictive value, NPV = 90%). Both false negative cases were upper GI bleed (vs 0 lower GI bleed); this difference was significant (p < 0.05). Conclusions The high NPV of CTA for the evaluation of GI bleeding suggests utility for excluding patients that are unlikely to benefit from MA and subsequent endovascular therapy. CTA may be considered for the first line diagnostic study for the evaluation of obscure GI bleeding.