Simulated colonoscopy training using a low-cost physical model improves responsiveness of surgery interns.

Abstract

Surgery residents are required to become proficient in colonoscopy before completing training. The aim of this study was to evaluate the responsiveness of surgery interns to simulated colonoscopy training. Interns, defined as postgraduate year 1 residents without exposure to endoscopy, underwent training in a physical model including colonoscopy, synthetic anatomy trays with luminal tattoos and a hybrid simulator. After baseline testing and mentored training, final testing was performed using five predetermined proficiency criteria. Content-valid metrics defined by the extent of departure from clinical reality were evaluated by two blinded assessors. Responsiveness was defined as change in performance over time and assessed comparing baseline testing with nonmentored final testing. Twelve interns (eight male, mean age 26, 80% right-handed) performed 48 colonoscopies each over 1year. Improvement was seen in the overall procedure time (24min 46s vs 20min 54s; P=0.03), passing the splenic flexure (20min 33s vs 10min 45s; P=0.007), passing the hepatic flexure (23min 31s vs 12min 45s; P=0.003), caecal intubation time (23min 38s vs 13min 26s; P=0.008), the duration of loss of view of the lumen (75% vs 8.3%; P=0.023), incomplete colonoscopy (100% vs 33.3%; P=0.042), colonoscope withdrawal <6min (16.7% vs 8.3%; P=0.052). Tattoo identification time (9min 16s vs 12min 25s; P=0.50), colon looped time (2min 12s vs 1min 45s; P=0.50) and rate of colon perforation (8.3% vs 8.3%; P=1) remained unchanged. Interrater reliability was 1.0 for all measures. Simulated colonoscopy training in a low-cost physical model improved the performance of surgery interns with decreased procedure time, increased rates of complete colonoscopy and appropriate scope withdrawal.