Capsule endoscopy enables noninvasive diagnostic examination of the entire small intestine. However, sensitivity and specificity of capsule endoscopy have not been adequately defined. We, therefore, compared capsule endoscopy by using intraoperative enteroscopy as a criterion standard in patients with obscure GI bleeding.

Forty-seven consecutive patients with obscure GI bleeding (11 with ongoing overt bleeding, 24 with previous overt bleeding, and 12 with obscure-occult bleeding) from two German gastroenterologic centers were included. All patients who had a prior nondiagnostic evaluation, including upper endoscopy, colonoscopy with a retrograde examination of the distal ileum, and push enteroscopy, underwent capsule endoscopy followed by intraoperative enteroscopy.

Capsule endoscopy identified lesions in 100% of the patients with ongoing overt bleeding, 67% of the patients with previous overt bleeding, and 67% of the patients with obscure-occult bleeding. Angiectasias were the most common source of bleeding (n=22). Capsule endoscopy showed the source of bleeding in 74.4% of all patients. The method was more effective in patients with ongoing bleeding. Compared with intraoperative enteroscopy sensitivity, specificity, and positive and negative predictive values of capsule endoscopy were 95%, 75%, 95%, and 86%, respectively.

Capsule endoscopy has high sensitivity and specificity to detect a bleeding source in patients with obscure GI bleeding. Thus, wireless capsule endoscopy can be recommended as part of the routine work-up in patients with obscure GI bleeding.