Background: The identification and treatment of lesions located in the small intestine in obscure gastrointestinal bleeding is always a clinical challenge.
Aim: To examine prospectively the diagnostic precision and the clinical efficacy of capsule endoscopy compared with push enteroscopy in obscure gastrointestinal bleeding.
Methods: Forty-two patients (22 men and 20 women) with obscure gastrointestinal bleeding (overt bleeding in 26 cases and occult blood loss with chronic anaemia in 16) and normal oesophagogastroduodenoscopy and colonoscopy were analysed. All patients were instructed to receive the capsule endoscopy and push enteroscopy was performed within the next 7 days. Both techniques were blindly performed by separate examiners. The diagnostic yield for each technique was defined as the frequency of detection of clinically relevant intestinal lesions carrying potential for bleeding.
Results: A bleeding site potentially related to gastrointestinal bleeding or evidence of active bleeding was identified in a greater proportion of patients using capsule endoscopy (74%; 31 of 42) than enteroscopy (19%; eight of 42) (P = 0.05). The most frequent capsule endoscopy findings were: angiodysplasia (45%), fresh blood (23%), jejunal ulcers (10%), ileal inflammatory mucosa (6%) and ileal tumour (6%). No additional intestinal diagnoses were made by enteroscopy. In seven patients (22%), the results obtained with capsule endoscopy led to a successful change in the therapeutic approach.
Conclusions: Compared with push enteroscopy, capsule endoscopy increases the diagnosis yield in patients with obscure gastrointestinal bleeding, and allows modification on therapy strategy in a remarkable proportion of patients.