What is the best test for evaluation of the small bowel in patients with gastrointestinal bleeding?

A strategy of capsule endoscopy (CE), followed by push enteroscopy (PE) only if the initial test is nondiagnostic, appears to be a less invasive, probably less expensive, and at least as sensitive as the opposite strategy. (LOE = 1b)

de Leusse A, Vahedi K, Edery J, et al. Capsule endoscopy or push enteroscopy for first-line exploration of obscure gastrointestinal bleeding? Gastroenterology 2007;132:855-862.  [PMID:17324401]

Randomized controlled trial (nonblinded)

Foundation

Concealed

Outpatient (specialty)

Obscure gastrointestinal bleeding is defined as bleeding whose source is not diagnosed after upper and lower endoscopy. In patients with iron deficiency anemia or overt bleeding, the source is often in the small bowel and PE is often performed. This invasive procedure requires sedation and is not able to visualize the entire small bowel. CE is an alternative that allows visualization of the entire small bowel and is less invasive than PE. In this study, patients with OGIB at 2 French medical centers were randomly assigned to initial CE or PE, followed by the alternative procedure if the initial study was nondiagnostic or if required for clinical reasons during follow-up. All patients had overt bleeding within the past 6 months or chronic iron deficiency anemia (> 3 months) and negative previous investigations (including upper endoscopy, colonoscopy, small bowel barium series, or computed tomographic enteroclysis). All patients younger than 30 years also had a Meckel scan. Thirty women and 48 men with a mean age of 54 years (range = 22 to 85 years) met inclusion criteria and were randomized; the onset of bleeding was a mean of 33 months ago. Three patients in the CE group and 4 in the PE group were lost to follow-up. After the initial study, a definite bleeding source was identified in 20 of 40 patients in the CE group compared with only 9 of 38 that underwent PE first (P = .02). The follow-up PE found 3 additional lesions in those who underwent CE first; conversely, the follow-up CE found additional lesions in 10 of 29 patients who had an initially negative PE. Thus, while the overall diagnostic yield was similar (58% for CE first vs 50% for PE first) fewer patients required the more invasive procedure if they had CE first. Also, although PE missed some lesions that were detected by CE, CE detected all lesions seen with PE.