Background: Wireless capsule endoscopy is a new method enabling non-invasive diagnostic endoscopy of the entire small intestine. In this study we prospectively examined the diagnostic precision of capsule endoscopy compared with push enteroscopy in patients with occult gastrointestinal bleeding.
Methods: Between July 2001 and October 2002 we examined 48 patients with suspected disorders of the small intestine using capsule endoscopy. 33 patients with obscure bleeding (19 men, 14 women, mean age 58 +/- 23 years) were prospectively examined using capsule endoscopy and push enteroscopy.
Results: On average, the patients had been suffering from chronic gastrointestinal bleeding for 30 +/- 36 (1-120) months. The lowest haemoglobin level was 6.5 +/- 1.6 g/dl (2.3-9.6) and on average 9 +/- 10 (0-50) blood units were transfused. Each patient underwent 4 +/- 2 (1-10) hospitalisations, with a mean 9 +/- 4 (5-17) diagnostic procedures before capsule endoscopy was used. Definitive bleeding sites were diagnosed by push enteroscopy in 7 patients (angiodysplasia [n = 5], ulcers [n = 1], multiple jejunal diverticula [n = 1]). Capsule endoscopy showed a bleeding source in 25 cases (76 %) (angiodysplasias [n = 15], Meckel's diverticulum [n = 1], ulcers [n = 7], ileum diverticulosis [n = 1], B-cell lymphoma [n = 1]). Push enteroscopy localised an additional bleeding source in comparison with capsule endoscopy (multiple jejunal diverticula) in one patient. Both methods of examination were safe and showed no complications.
Discussion: The present study shows that capsule endoscopy had the highest diagnostic yield and was superior to push enteroscopy in patients with chronic gastrointestinal bleeding. By using the capsule at an early stage the subsequent therapeutic procedure could be considerably shortened and diagnostic processes could possibly be optimised.