Prospective controlled assessment of variable stiffness enteroscopy

Background: Push enteroscopy is a well-established technique for evaluation of the small intestine. However, looping of the enteroscope within the stomach limits depth of insertion. Stiffening overtubes that minimize gastric looping are tolerated marginally by patients and disliked by endoscopists. A variable stiffness instrument has the potential to eliminate the need for an overtube while still minimizing gastric looping. The performance of a prototype variable stiffness enteroscope was compared prospectively with that of a conventional push enteroscope with and without use of an overtube. Methods: Consecutive patients undergoing enteroscopy were randomized to have the procedure with a variable stiffness instrument, a conventional instrument with overtube, or a conventional instrument without overtube. Depth of insertion distal to the ligament of Treitz was determined by plain abdominal radiography. Observations: In total, 67 patients were randomized to variable stiffness enteroscopy (25 patients), enteroscopy with overtube (23 patients), and enteroscopy without overtube (19 patients). Median depth of insertion distal to the ligament of Treitz, respectively, for each group, was 89 cm, 68 cm and 41 cm (p = 0.03). In multivariate analysis, variable stiffness instrument use was predictive of intubation to 65 cm distal to the ligament of Treitz (odds ratio 5.53: 95% Cl [1.25, 31.25] vs. no overtube, and odds ratio 2.50: 95% Cl [0.63, 11.1] vs. overtube). Procedure duration and overall patient tolerance did not differ significantly among the 3 groups, although more patients in the overtube group required additional sedation than patients in the variable stiffness group (p = 0.03). Both endoscopists' (r = 0.34) and nurses' (r = 0.36) estimates of patient discomfort during the procedure correlated poorly with patient tolerance. Conclusions: A variable stiffness push enteroscope enhances insertion depth compared with the conventional instrument with or without overtube. Further studies are required to determine whether this improved performance increases diagnostic yield.