Simulation for ERCP training: A good idea that refuses to die …

Endoscopic simulation is like the prize fighter in the Rocky movies: it gets knocked down repeatedly and almost counted out, but somehow manages to stagger to its feet – bloodied, but with its head held high – to fight another round.

I have been an observer of, and sometime participant in, endoscopic simulation training for over 25 years [1] [2] [3]. It has been a bittersweet experience. Something we early investigators thought would be a “no brainer” – developing realistic simulations of endoscopic procedures – turned out to be anything but simple. The desktop computers of the 1980 s seemed perfect for the purpose. Video games were all the rage: surely the computer geeks who brought us PacMan and Donkey Kong could whip up a cheap-and-cheerful interactive simulation of the luminal gut? Presented with the basic outline of a computerized colonoscopy simulator, one of the major video game makers told me it could be done, but that it would take 20 of his programmers working for 6 months at a cost of several million US dollars. And the input device (the dummy colonoscope) would have to be provided to them, unless we had another couple of million dollars to spare…

One thing that united the small band of independent US and European simulation investigators was lack of money: we quickly discovered that the professional societies and governmental funding sources had no category that recognized our endeavors. We knocked on a lot of doors looking for research funds. Surprisingly few projects were funded by the major endoscope manufacturers. It was great fun developing computer simulation models, but it quickly became apparent that we had chosen a huge mountain to climb. The desktop computers of the 1980 s were hard pressed to deliver 16 or more frames per second of graphics, the minimum needed to create a realistic moving image. The introduction of the Intel 286 then 386 processors was a quantum leap at the time, offering hope that one day speed would no longer be rate-limiting: endoscopy simulation was seemingly insatiable, consuming vast amounts of memory and processing power. Despite solid progress, it was clear after years of work that only major funding would render realistic, accessible computer simulation available to the training masses. In the early 1990 s a consensus conference was held in Munich – coincidentally, home to the journal Endoscopy – to bring the various research groups together. We had high hopes of establishing our credentials and securing meaningful funding from industry for the next decade of development and beyond. Sadly, we could not convince industry that simulation training had a future, and what funding there had been quickly evaporated.

As computer simulation for endoscopic training languished, other investigators went back to simple mechanical models. After all, one of the earliest simulators – for flexible sigmoidoscopy – was constructed from the hose of a vacuum cleaner, twisted into shape and pinned on a board! These static simulators were good for demonstrating one or two basic “tricks,” like demonstrating what endoscope control inputs do, but they lacked realism and were quickly abandoned by trainees anxious to get doing actual procedures in real patients. Other investigators, mainly European, developed in vivo and ex vivo animal models for endoscopic training, with considerable success [4] [5]. Such models are now regularly used at training courses to offer hands-on experience of everything from endoscopic clip application to NOTES (natural-orifice transluminal endoscopic surgery) procedures. Useful as they are, ex vivo models require considerable preparation of preserved animal organs, usually by a dedicated technician. In vivo models are even more realistic, but also more expensive, and come with ethical concerns about using live animals for this activity.

Professor Joseph Leung and his colleagues at the University of California (Davis) have returned to basics to simulate endoscopic retrograde cholangiopancreatography (ERCP). Combining low tech (a basic mechanical model) with inexpensive high(er) tech (e. g., video cameras to simulate fluoroscopy), they have created an ERCP simulation for both training and evaluation [6]. As evidenced by their prior publications, they appear to have hit on an ERCP simulation that is sufficiently realistic to maintain their trainees’ attention, while inexpensive enough to be made widely available when the time comes (the Leung ERCP simulator remains a research tool, and is not yet available for purchase).

A major criticism of endoscopy simulation in the past has been that it is “too easy,” and therefore not a meaningful test of trainee skills. Studying simulators in a rigorous scientific fashion is difficult, resulting in a paucity of meaningful published data. Leung et al. have taken a rigorous scientific approach to the development and evaluation of their static ERCP simulator that is producing results. In this issue of Endoscopy, they report a study comparing long- and short-wire technology for placing biliary stents [7]. They show that the performance of a large group of trainees using the two different technologies can be evaluated objectively using their simulator model. They also show that, while there was no difference in the time taken to place a single biliary stent using long- and short-wire systems, the system that allowed intraductal release of the guide wire was superior for placing multiple (3) stents during the same procedure. As an extra bonus, the trainees got to balloon dilate a simulated biliary stricture before stenting it.

This study may seem like one small step for endoscopy training, but in many ways it represents one giant leap for ERCP simulation. The ability to compare endoscopic accessories in the simulation environment has been top of industry’s “wish list” for years. Taking the patient even partially out of the evaluation process may greatly simplify the expensive regulatory “hoops” that equipment makers currently have to go through. Reproducible comparisons between accessories using inexpensive simulators will also minimize the guess work and advertising hype. Just as Rome was not built in a day, Dr Leung’s ERCP simulator continues to be developed and refined. It is gratifying to see endoscopy simulation research supported by one of our professional organizations (Dr Leung received a 2007 Research Award from the American College of Gastroenterology) and a number of industry partners.

Over the years, I have heard endoscopy simulation “trashed” on many occasions by self-appointed experts pontificating on the subject. Like Rocky, simulation training has shown a commendable ability to get back on its feet after such “knockdowns” and keep on going [8]. Gastrointestinal endoscopy lags years behind other procedural specialties in medicine in its appreciation of the benefits of simulation training and evaluation. As Dr Leung sagely observed recently, “Perhaps the biggest problem is that (our generation) grew up learning procedures on patients.” He is right. The YouTube, video gamer, techno generation of trainees doesn’t have that barrier to learning in a simulation environment. So, the “old folk” of gastrointestinal endoscopy need to take a dance partner in the “simulation waltz,” or go sit this one out. Endoscopy simulation is getting up, dusting itself off, and moving on!

Competing interests: None

References

• 1 Baillie J, Jowell P, Evangelou H. et al . Teaching by endoscopy simulation.  Endoscopy. 1991;  23 239-40
• 2 Baillie J. Endoscopic retrograde cholangiopancreatography simulation.  Gastrointest Endosc Clin N Am. 2006;  16 529-542
• 3 Testoni P A, Sultan S, Baillie J. Can computer simulation accelerate the development of procedural competence?.  Am J Gastroenterol. 2004;  99 38-39
• 4 Neumann M, Mayer G, Ell C. et al . The Erlangen Endo-Trainer: life-like simulation for diagnostic and interventional endoscopic retrograde cholangiography.  Endoscopy. 2000;  32 906-910
• 5 Matthes K, Cohen J. The Neo-Papilla: a new modification of porcine ex vivo simulators for ERC P training.  Gastrointest Endosc. 2006;  64 570-576
• 6 Leung J W, Lee J, Rojany M. et al . Development of a novel ERCP mechanical simulator.  Gastrointest Endosc. 2007;  65 1056-1062
• 7 Leung J W, Lee W, Lim B S, Leung F W. Comparison of accessory performance using a novel ERCP mechanical simulator.  Endoscopy. 2008;  40 983-988
• 8 Sedlack R, Petersen B, Binmoeller K, Kolars J. A direct comparison of ERCP teaching models.  Gastrointest Endosc. 2003;  57 886-890