Endoscopic Clip-Assisted Feeding Tube Placement Reduces Repeat Endoscopy Rate: Results from a Randomized Controlled Trial

OBJECTIVES: To determine whether endoscopic clip-assisted nasoenteral feeding tube placement is more effective than standard feeding tube placement with transnasal endoscopy. METHODS: Between August 2009 and February 2011, 143 patients referred for endoscopic nasoenteral feeding tube placement were randomized between clip-assisted and standard nasoenteral tube placement. Endoscopies were performed in the endoscopy unit and intensive care unit in a tertiary referral center in the Netherlands. For the clip-assisted procedure, the feeding tube was introduced with a suture fixed to the tip, picked up in the stomach with an endoclip and attached (as distal as possible) to the duodenal wall. In the standard group, a guide wire was placed in the duodenum using a transnasal endoscope, followed by blind insertion of a feeding tube over the guide wire. Primary end point was a repeat endoscopy for incorrect tube placement or spontaneous retrograde tube migration. Secondary end points were incorrect tube placement, spontaneous migration of feeding tube, directs medical costs, and procedure-related (serious) adverse event (SAE). RESULTS: Of the 143 patients included, 71 were randomly assigned to clip-assisted tube placement, and 72 to standard tube placement. Four (5.6%) repeat endoscopies were performed in the clip-assisted group vs. 19 (26.4%) in the standard group (relative risk reduction (RRR) 0.79; 95% confidence interval (CI) 0.40-0.92). The number needed to clip to avoid one repeat endoscopy was 4.8 (95% CI 3.1-11.3). Repeat endoscopies were mostly performed for incorrectly placed tubes, 3 (4.2%) in the clip-assisted group vs. 16 (22.2%, RRR 0.81; 95% CI 0.38-0.94) in the standard group. Spontaneous retrograde tube migration occurred in one (1.4%) clip-assisted placement and three (4.2%) standard tubes. Median costs were higher for clip-assisted tube placement ((sic)519 vs. (sic)423, P < 0.01). Four (5.6%) SAEs occurred after clip-assisted feeding tube placement vs. one (1.4%) after standard feeding tube placement (P = 0.21). CONCLUSION: Clip-assisted endoscopic nasoenteral feeding tube placement results in fewer repeat endoscopies than standard endoscopic nasoenteral tube placement, due to a higher success rate of initial placement. When tubes are adequately placed, retrograde tube migration rarely occurs.