Endoscopic ultrasound teaching and learning.

Scientific publications in the field of endoscopic ultrasound (EUS) are raising up since the last twenty years. The value of EUS is directly proportional to the training, skill and experience of the endosonographer. In a recent series, the lowest annual number of examinations requested per year might be around 200-250 procedures. For achieving competence in all the fields of EUS, they recommended a minimum of 150 supervised cases, 75 of them being devoted to pancreaticobiliary diseases and 50 to fine needle aspiration (FNA). Many studies have yet confirmed the importance of the learning curve to improve the EUS accuracy. The accuracy of EUS staging of esophageal cancer was significantly improved over an 8-year period for T staging from 64% to 90% but not for N staging. Considering pancreaticobiliary diseases, some series reported the influence of training on accuracy of EUS-guided FNA of pancreatic masses. Even if the exact type and modalities for EUS training remained to be determined, the usefulness of a period of training in a center performing more than 200 procedures per year is confirmed. Improving cognitive EUS learning and technique required theoretical teaching and practical teaching with observation during live demonstration and echoendoscope hands-on in humans with an expert supervision or with different simulator models. In France, formal EUS training has been underway since 1991, with a two-year intensive program including 4 weeks of theoretical training and 20 one-day practical sessions per year. This program was updated in 2004 to include 3 weeks of theoretical course work including numerous video sessions and 1 week of practical sessions. Theoretical courses are given primarily by gastroenterologists and EUS experts, but also by pathologists, surgeons, anatomists and radiologists. Practical sessions include 4 days of live demonstrations and computed-based simulator (EUS mentor; Simbionix, Tel Hashomer, Israel) and one day with a swine model. Simulators, either computed-based simulator, Erlangen model, or live pigs are believed to shorten the learning phase of EUS. The live pig model was chosen for EUS credentialing because it appears to be similar in many respects to the human anatomy, especially for the pancreas, portal and mesenteric vessels, celiac axis, left kidney and spleen. In addition, for trainees, it often appears more exciting and stimulating than virtual model. However, EUS training on a swine model is recognized to be the more expansive way for EUS teaching. Between the pre and post-test, EUS fellows significantly improved their competence for the visualization of anatomical structures i.e. vena cava, mesenteric/splenic vein, celiac axis, pancreas and bile duct. At the end of the day-session, trainees were mostly able to recognize and follow these anatomical structures with both endoscopes. Interventional EUS learning was also assessed. A significant decrease in procedural duration and a significant increase in FNA precision were demonstrated for the puncture of a hilar liver lymph node.