Is Small-Bowel Radiography Necessary Before Double-Balloon Endoscopy?

Abstract

Small-bowel radiography may be replaced by enteroscopy in the diagnosis of small-intestine lesions. We retrospectively elucidated the diagnostic yield of small-bowel radiography performed before double-balloon endoscopy. One hundred twenty-four patients who underwent double-balloon endoscopy during the period 2004-2006 were classified into those with abnormal radiographic findings (n = 45), normal radiographic findings (n = 31), and no small-bowl radiographs (n = 48). The classification was based on the use of small-bowel radiography and the diagnosis before double-balloon endoscopy. The indications for, approaches to, and diagnostic yields of double-balloon endoscopy were compared for the three groups. The diagnostic yield of small-bowel radiography was considered positive when any sign of pathologic change in the small bowel was identified. The diagnostic yield of double-balloon endoscopy was considered positive when endoscopic or biopsy findings explained the clinical manifestations. The group with abnormal findings on small-bowel radiography was younger (15-86 years) and less frequently had obscure bleeding (8.9%) than the group with normal findings on small-bowel radiography (age, 17-84 years; frequency of obscure bleeding, 45.2%) (p = 0.01) or the group without small-bowel radiographs (age, 15-91 years; frequency of obscure bleeding, 64.6%) (p < 0.0001). The positive diagnostic yield of double-balloon endoscopy was highest in the group with abnormal findings on small-bowel radiography (71.1%), followed by the group with no small-bowel radiographs (45.8%) and the group with normal findings on small-bowel radiography (35.5%) (p = 0.0002). Among patients who did undergo small-bowl radiography, the accuracy of the technique was 68.4%, the positive predictive value was 71.1%, and the negative predictive value was 64.5%. The positive diagnostic yields of small-bowel radiography and double-balloon endoscopy were not statistically different (59.2% for small-bowel radiography, 56.6% for double-balloon endoscopy; p > 0.1). The diagnostic accuracy of double-balloon endoscopy seems to improve if the procedure is preceded by small-bowel radiography.