Prospective Blinded Comparison of Wireless Capsule Endoscopy and Multiphase CT Enterography in Obscure Gastrointestinal Bleeding

Abstract

To compare the performance of multiphase computed tomographic (CT) enterography with that of capsule endoscopy in a group of patients with obscure gastrointestinal bleeding (OGIB). This prospective HIPAA-compliant study was approved by the institutional review board and the institutional conflict of interest committee. All patients provided written informed consent. Two radiologists, blinded to clinical data and results of capsule endoscopy, interpreted images from CT enterography independently, with discordant interpretations resolved by consensus. Results were compared with those from a reference standard (surgery or endoscopy) and clinical follow-up. Sensitivity and 95% confidence intervals were calculated for each modality. Fifty-eight adult patients, referred for the evaluation of OGIB (occult, 25 patients [43%]; overt, 33 patients [57%]), underwent both tests. A small bowel bleeding source was identified in 16 of the 58 patients (28%). The sensitivity of CT enterography was significantly greater than that of capsule endoscopy (88% [14 of 16 patients] vs 38% [six of 16 patients], respectively; P = .008), largely because it depicted more small bowel masses (100% [nine of nine patients] vs 33% [three of nine patients], respectively; P = .03). No additional small bowel tumors were discovered during the follow-up period (range, 5.6-45.9 months; mean, 16.6 months). In this referral population, the sensitivity of CT enterography for detecting small bowel bleeding sources and small bowel masses was significantly greater than that of capsule endoscopy. On the basis of these findings, the addition of multiphase CT enterography to the routine diagnostic work-up of patients with OGIB should be considered, particularly in patients with negative findings at capsule endoscopy.