Capsule endoscopy versus computed tomographic or standard angiography for the diagnosis of obscure gastrointestinal bleeding.

BACKGROUND AND AIMS: Capsule endoscopy (CE) is superior to push enteroscopy and small bowel barium radiography in detecting the source of obscure GI bleeding. We now compared whether CE has a superior diagnostic yield than CT angiography (CTA) or standard mesenteric angiography (ANGIO) in patients with obscure GI bleeding. METHODS: From June 2004 to October 2005, consecutive patients admitted for OGIB underwent both CTA and ANGIO, followed by CE, performed blindly by independent examiners within the next 7 days. The primary end point of the study was the diagnostic yield for each technique, defined as the frequency of detection of lesions with a high probability of bleeding. RESULTS: Twenty-eight patients (16 men and 12 women, mean age 74 +/- 2 yr) with OGIB (overt bleeding in 20 cases and chronic occult in 8) were prospectively evaluated. CTA or standard angiography could be performed in 25 of 28 patients (applicability 86%), because of contrast allergy (1 patient) and chronic renal failure (2 patients). A source of bleeding was detected by CE in a greater proportion of patients, 72% (18 of 25, 95% CI 50.6-87.9%), than CTA, 24% (6 of 25, 95% CI 9.4-45.1%, P= 0.005 vs CE), or ANGIO, 56% (14 of 25, 95% CI 34.9-75.6%, P= NS). Similarly, CE was able to diagnose 100% of patients diagnosed by CTA and 86% of patients diagnosed by ANGIO. Moreover, CE was positive in 12 of 19 (63%) negative cases on CTA and in 6 of 11 (55%) negative cases on ANGIO. As a result of the CE findings, therapeutic intervention was undertaken in 9 of 19 (47%) patients with positive results. CONCLUSION: CE detects more lesions than CTA or standard mesenteric angiography in patients with obscure GI bleeding and has a therapeutic impact in almost half of the patients with positive findings.