Search

To retrospectively evaluate the performance of computed tomography (CT) angiography in the detection and localization of clinically active gastrointestinal (GI) hemorrhage of an unknown source.Eighty-six CT angiograms were obtained in 74 patients with the clinical diagnosis of acute GI hemorrhage of an unknown source. Results of CT angiography were recorded, and the patients' electronic medical records were reviewed for documentation of subsequent interventional procedures performed within 24 hours of the reference CT angiogram to diagnose or control ongoing GI hemorrhage. Surgical, endoscopic, and final pathologic reports, if available, were reviewed.Twenty-two of the 86 CT angiograms (26%) were positive for active hemorrhage, with findings confirmed in 19 of the 22 cases (86%). Thirteen cases were confirmed with angiography, five cases were confirmed with surgery, and one case was confirmed with autopsy. Sixty-four of the 86 CT angiograms were negative, and 59 (92%) of the CT angiograms required no further intervention. These patients were discharged without incident. There were no cases in which CT angiography was negative and subsequent angiography within 24 hours was positive. The overall sensitivity, specificity, accuracy, and positive and negative predictive value of CT angiography in the detection of active GI hemorrhage within this study population were 79%, 95%, 91%, 86%, and 92%, respectively.CT angiography provides valuable information that can be used to determine the appropriateness of catheter angiography and guide mesenteric catheterization if a bleeding source is localized. The authors' experience with this study cohort supports its use before angiography in those patients with acute GI bleeding of an unknown source who are being considered for catheter-directed intervention.