Background and study aims: In chronic gastrointestinal bleeding (CGB), bleeding sites located in the small bowel are difficult to detect with conventional radiological or scintigraphic techniques. Push enteroscopy (PE) is at present considered to be the most effective diagnostic procedure. The aim of this prospective trial was to compare the efficacy of wireless capsule endoscopy (CE) with PE.
Patients and methods: Between April and October 2001, 65 patients with CGB were referred to our unit. Complete conventional diagnostic work-up (including small-bowel enteroclysis, angiography, and scintigraphy), as well as PE and CE, were performed in 32 patients.
Results: On average, the patients had been suffering from CGB for 29 +/- 24 months (6 - 126); the lowest hemoglobin level varied between 3.0 and 9.9 g/dl (mean 5.9 +/- 1.4); 17 +/- 18 blood units (0 - 60) were transfused. Each patient underwent 6 +/- 7 (range 1 - 38) hospitalizations, with a mean of 14 +/- 9 diagnostic procedures before CE was used. Conventional diagnostic procedures revealed relevant pathological findings in five of the 32 patients (16 %). Definite bleeding sites diagnosed by PE in nine patients (28 %) included angiodysplasia (seven patients), small-bowel cancer (one patient) and lymphoma (one patient). CE detected the definite source in 21 of the 32 patients (66 %) ( P < 0.001). Definite bleeding sources included angiodysplasia (17 patients), malignant stenoses (two patients) and inflammatory small-bowel disease (two patients). Questionable bleeding sources were seen on PE in three additional patients (9 %), and using CE in a further seven patients (22 %). Both CE and PE were safe and were not associated with any morbidity.
Conclusions: In the present trial in patients with CGB, wireless CE had the highest diagnostic yield and was significantly superior to PE. CE can help reduce the number of diagnostic procedures and could become the initial diagnostic choice in patients with CGB and negative upper and lower gastrointestinal endoscopy.