Gastroenterological Endoscopy. Группа авторов

activities cannot replace the performance of the actual procedures.

      1.3 Incorporation of Simulators in Training

      The Gastroenterology Core Curriculum, Third Edition in May 2007 states in section IV.A.6.(b): “Fellows must participate in training using simulation.”²³ To date, no simulator experience alone has been validated as sufficient to replace actual patient experience. To guide adoption of simulators for specific roles in training and assessing skill, the ASGE initiated a PIVI (Preservation and Incorporation of Valuable Endoscopic Innovations) task force in 2011.²⁸ This group set the following two thresholds for justifying adoption of a particular simulator:

      • Threshold for incorporation into training. For an endoscopy simulator to be integrated into the standard instruction for a procedure, it must demonstrate a 25% or greater reduction in the median number of clinical cases required for the trainees to achieve the minimal competence parameters for that procedure.

      • Threshold for assessing skill. Simulator-based assessment tools must be procedure-specific and predictive of independently defined minimal competence parameters from real procedures with a kappa value of at least 0.70 for high-stakes assessment.²⁸

      The logistic and cost issues for a particular simulator would need to be weighed. For example, a high-cost computer simulator that had a 25% reduction in a learning curve might not make any sense for a program in which trainees typically had sufficient actual case experience to develop competency. In contrast, a lower cost simulator in which a program typically had insufficient cases would be well worth the investment.

      1.4 Endoscopy Simulators and Training Models

      1.4.1 Plastic Phantoms and Other Static Models

      The initial experimental models for endoscopy training were made of plastic and textile tissues.¹ In 1974, Classen and Ruppin²⁹ in Erlangen presented an anatomically shaped plastic phantom that allowed examination of the upper GI tract. Christopher Williams and his group in London have been working on the first semi-rigid colonoscopy phantoms. A robust further development represents the Kyoto Kagaku Colonoscope Training Model, which presents greater technical difficulty to reach the cecum and allows a more realistic loop reduction

Fig. 1.6.²⁹ Grund and co-workers in Tübingen, Germany, developed a series of advanced static models for different training purposes.^32,33 They include artificial tissues for electrosurgical interventions and recently specific ERCP techniques. Unfortunately, those models are not commercially available so far and there are no published data validating their use in training.

      In addition, a number of device manufacturers have produced their own models to facilitate training in the procedures in which their accessories are used. The Cook Medical ERCP Trainer recently developed by Costamagna et al³⁴ allows to practice cannulation and different ERCP techniques except sphincterotomy via a plastic papilla with varying ampullary anatomy, orientation, and cannulation difficulty.

      Another promising simulator is the “T.E.S.T box simulator” (

Fig. 1.5).³⁵ The model, designed by Christopher Thompson has demonstrated an ability to distinguish skills levels with significant differences between all categories from beginner to expert interventional endoscopist. One limitation of this and all static simulators to date is the limited exposure to pathology for training in image recognition and application of findings into management decisions.

      Fig. 1.5 The Thompson Endoscopic Part Task Simulator Training (T.E.S.T.) box containing five different training modules.³³

      1.4.2. Computer Simulators

      Various computer simulation systems have been developed since the early 1980s.¹ Rapid progress in computer technology and electronics at the early 2000s allowed the development of commercially available systems. The first of these models was the Simbionix GI Mentor (3D Systems Healthcare, Littleton, CO, United States, formerly Simbionix Corporation), at the time in the shape of a human torso mannequin.³⁴ The system creates a relatively realistic virtual endoscopy environment and allowed the simulation of various diagnostic and interventional procedures at different levels. During training, teaching modules with anatomy and pathology (

Fig. 1.7) atlases are at the trainee’s disposition. Beginners can train their dexterity in a “GI Fundamental Skills” module including navigation, targeting, retroflection, loop reduction, or in “Cyberscopy,” a module to further enhance hand–eye coordination. Different modules such as upper and lower GI endoscopy, sigmoidoscopy, EUS, ERCP, and hemostasis training are available. EUS and ERCP modules allow parallel viewing of radiographic and endoscopic simulations. Virtual sphincterotomy, stone extraction, and other techniques have been implemented. In addition to the current GI Mentor model (3D Systems Healthcare), the EndoVR virtual reality endoscopy simulator (CAE Healthcare, Montreal, Canada, formerly “Accutouch” by Immersion Medical, Inc., Gaithersburg, MD, United States) has been used in multiple studies (see later). Recently, another system the so-called “Endo X” has been presented (Medical-X BV, EM Rotterdam, the Netherlands (

Fig. 1.8). The system provides mainly upper and lower GI techniques, but also includes analyzing tools such as insufflation performance simulation and video recording of the procedure. All devices allow user-specific training curricula and reflect the user-specific learning curve. Modules are supervised by a virtual tutor and the whole system can be connected to a real supervisor via internet for additional personal feedback and to view learning curves of different trainees by the supervisor (

Fig. 1.9). Various studies have demonstrated the benefits of additional computer simulator training in connection with colonoscopy.^1,28

      In a prospective simulation study, four fellows at the Mayo Clinic received 6 hours of simulator-based training, compared with four fellows without training. The simulator-trained fellows outperformed the traditionally trained fellows during their initial 15 to 30 colonoscopies in all performance aspects except for insertion time (p < 0.05). Beyond 30 procedures, there were no differences in performance between the two groups (evidence level B).

      Fig. 1.6 The Kyoto Kagaku colonoscopy training model with different possibilities to vary the difficulty of passage of the sigmoid (Level 1-6). (Images are provided courtesy of Kyoto Kagaku, Kyoto, Japan.)

      Fig. 1.7 Hands-on training using the compactEASIE simulator. (a) Groups of three or four fellows per simulator and teacher receiving instructions. (b) Individual practice, for example, for basic gastroscopy.¹

      Fig. 1.8 The EASIE-R model designed by Kai Matthes and based on the compactEASIE