A second-generation virtual reality simulator for colonoscopy: validation and initial experience.

BACKGROUND AND STUDY AIMS: Simulators are increasingly used in skills training for physicians; however data on systematic evaluation of the performance of these simulators are scarce compared with those used in aviation. The objectives of this study were to determine the expert validity, the construct validity, and the training value of the novel Olympus simulator as judged by experts. PATIENTS AND METHODS: Participants were novices and experts. Novices had no prior experience in flexible endoscopy; experts had all performed more than 1000 colonoscopies. Participants filled out a questionnaire on their impression of the realism of the colonoscopy exercises performed. These included a dexterity exercise and a virtual colonoscopy. Test parameters used were points acquired in a game, time to reach the cecum, maximum insertion force, and "patient pain." RESULTS: Novices (n = 26) scored a median of 973 points (range--118-1393), experts (n = 23) scored 1212 points (range 89-1375). This difference did not reach significance (P = 0.073). Experts performed virtual colonoscopy significantly faster than novices (220 vs. 780 s, P < 0.001) but used more insertion force (11.8 vs. 11.6 N; P = 0.147). Maximum pain score was higher in the expert group: 86% vs. 73%. (P = 0.018). The realism was graded 6.5 on a 10-point scale. Experts considered the Olympus simulator beneficial for the training of novice endoscopists. CONCLUSIONS: The novel Olympus simulator discriminates excellently between the measured levels of expertise. The prototype offers a good realistic representation of colonoscopy according to experts. Although the software development is continuing, the device can already be implemented in the training program of novice endoscopists.