Reproducibility of wireless capsule endoscopy in the investigation of chronic obscure gastrointestinal bleeding.

BACKGROUND Capsule endoscopy (CE) is a valuable tool in the diagnostic evaluation of obscure gastrointestinal bleeding, but limited information is available on the reproducibility of CE findings. OBJECTIVE To compare two successive CE studies with push enteroscopy (PE) in patients presenting with chronic obscure gastrointestinal bleeding. METHODS A prospective study was conducted. Ten patients (seven men and three women) with chronic obscure gastrointestinal bleeding and no contraindications for CE were eligible and completed the trial. For each patient, the first capsule was administered on day 1, the second capsule was administered on day 2 and PE was performed on day 3. Endoscopists were blinded to the capsule findings. Capsule findings were assessed independently by two investigators blinded to PE findings. RESULTS A potential small intestinal bleeding source was found in 60% of the patients when all the studies were combined. A bleeding source was found in four patients in both CE studies. The second CE also identified a bleeding source in a fifth patient. Interobserver agreement by kappa analysis was 0.642 to 1.000 (P < or 05) for the CE studies. PE identified a potential small bowel bleeding site in four patients, including one patient who had negative CE studies. CONCLUSIONS This study confirmed the reproducibility of CE findings on successive studies. Some patients did not have a source of bleeding in the small intestine, and all studies found this.