Synopsis  

Training in gastrointestinal endoscopy represents an ever-increasingly important endeavor as procedures have become more complex and more therapeutic in nature. In the US, the bulk of this instruction now occurs during formalized training programs of at least 3 years duration. Additional periods of dedicated intensive training in certain specialized procedures such as ERCP and EUS are available, although many practitioners routinely perform such procedures without added formal training. In addition to the supervised instruction of novices during fellowship, training also encompasses the activities of experienced practicing gastroenterologists trying to learn new techniques or refresh certain skills that have become rusty from disuse.

The goal of all training is to become proficient in the techniques being taught. A lot of effort is being made to define when someone is truly competent to perform a procedure independently and to determine how much training is required to reach this level of skill. In the past, the number of cases performed under supervision has sufficed as a surrogate for actual demonstration of competency. A wide range of minimum case numbers have been recommended for different procedures, often based on expert opinion. More recently, investigators have attempted to correlate the number of cases performed with objective criteria for success. Increasingly, information about the amount of training required to master a given procedure and the practice of objective competency assessment will impact the way in which hospitals grant individuals privileges to perform endoscopic procedures.

Novel methods of instruction have been introduced to accelerate the learning curve of certain techniques at various stages in the training process. Endoscopy simulators, including ex vivo animal tissue, artificial tissue, and virtual reality computer-based models, have exciting potential to enhance the existing practice of teaching endoscopy. In addition, these simulators may provide a way to assess competency in a controlled environment that does not affect patient care and could one day be used to facilitate the process of credentialing and recredentialing for endoscopy privileges. However, the precise role and the optimal use of the various models have not yet been clearly defined, and the presumed benefit remains to be confirmed by controlled studies.

This chapter will first review general principles of endoscopy training and the specific skill sets required to gain proficiency. It will then address each of the major endoscopic procedures, the type and amount of training currently recommended to master each procedure, and the evidence which supports current recommendations. Criteria used to assess competency for each procedure will be discussed. This chapter will then address the role of a wide variety of alternative methods for teaching endoscopy, which are available to supplement standard training. Finally, current recommendations for hospital credentialing and privileging will be discussed.

 General principles of endoscopy training   

Traditional standard means of instruction   

Hands-on, supervised one-on-one instruction is the mainstay of endoscopy training. Successful teaching with this method depends on many factors.

Teachers   

Perhaps the most important ingredient to effective teaching is the teaching skill of the endoscopy instructor and the quality of the communication that takes place between trainer and trainee [1]. Rather than just watch a trainee attempt a procedure and take over when they experience difficulty or when they have used up a predetermined period of time, the instructor is supposed to engage in a dynamic process of mostly verbal description, feedback, and inquiry with the trainee during each procedure. This is often difficult to do, even for expert endoscopists who must convey verbal instructions about maneuvers that they routinely perform without much thought. It requires the ability to break down the technical components into discrete steps for the novice. Just as it is difficult to tell someone how to tie a shoelace using only words, the instructor combines verbal instruction with brief demonstrations to illustrate a particular technique. Other important skills required of the endoscopy teacher include the patience to allow the trainee time to develop skills without endangering patient safety, and the ability to give constructive feedback [2,3].

Environment   

Apart from the quality of the instructor, optimal training conditions also depend upon a relaxed learning environment in the endoscopy unit with logistics that allow for the added time necessary for fellows to perform procedures. The training setting must also provide trainees with sufficient case volume for adequate repetition of skills, and a wide enough variety of pathology to allow them to develop cognitive assessment skills in parallel with technical expertise.

Is self-teaching still acceptable?   

This supervised 'hands-on' training method has generally been an effective means of developing proficiency in a wide variety of procedures. It has largely supplanted the 'self-taught' method by which many endoscopy pioneers learned and mastered endoscopic techniques before expert instruction became widely available. While some practitioners do attempt to perform new techniques with only perfunctory supervised hands-on instruction, this practice opens up a wide array of ethical and practical concerns.

Can one justify 'see one, do one' if expert supervised training is available as an alternative? For specific new techniques such as endoscopic suturing, radiofrequency ablation of the lower esophageal sphincter (LES), or metal stent placement, can expert endoscopists just begin performing procedures after attending a short course, seeing some demonstrations, and watching a videotape presentation? Such practice may seem questionable from an ethical standpoint when there are feasible alternative methods available to learn new procedures for individuals already in practice. However, are there sufficient opportunities available for practitioners to receive sufficient supervised formal training to develop the necessary skills prior to using new techniques on their own? A perceived need for such opportunities led to a trial American Society for Gastrointestinal Endoscopy (ASGE) preceptorship program described in 1992 by Borland, but this pilot effort did not take hold [4]. The ASGE guidelines concerning so-called 'short courses' suggest a potential role of such venues for experienced endoscopists to learn new techniques, while rejecting this method for alternative primary training for major endoscopic procedures such as colonoscopy, ERCP, EUS, photodynamic therapy, and laser therapy [5].

What to teach and how to teach it   

What skill sets must a trainee acquire in order to learn to perform endoscopic procedures? The novice endoscopist must develop both technical and cognitive proficiencies in the following areas:

• Understanding the indications and contraindications for endoscopic procedures, and risk factors for complications.
  
• Knowledge of the endoscopic equipment and accessories and how to set up this equipment for use.
  
• Familiarity with the endoscope control dials and buttons.
  
• Dexterity in controlling the scope range of motion using the dials and torque applied to the endoscope shaft.
  
• Hand–eye coordination to produce deliberate, precise manipulation of the scope within the lumen, and to use accessories.
  
• Communication with nursing and technical staff regarding required assistance during the procedure.
  
• Knowledge of normal anatomical landmarks and possible abnormal pathologies which might be encountered.
  
• Interpretive skills to correctly identify abnormalities which are detected.
  
• Judgement of how to manage appropriately those lesions encountered.
  
• Familiarity with patient monitoring and the administration of conscious sedation.
  
• Awareness of how to recognize and manage adverse events.
  
• Understanding of risks and benefits of intended procedures and the ability to obtain informed consent.
  
• Documentation of findings.
  
• Communication of results to patients and other physicians.
  

While many of these activities seem obvious, they are all critical skills which the endoscopic trainee must master before performing procedures independently. Some of this material may be included in a formal curriculum designed for the novice as an introduction to endoscopy. However, much is introduced to the new fellow more haphazardly as he or she begins to perform cases under supervision by one or more preceptors. While there has been increased focus within training programs on objective competency assessment, evaluation of fellows' progress is seldom broken down into such a detailed list of skill sets.

Defining competency and how to access it   

Before determining how best to train fellows, it is necessary first to establish what constitutes the end product of successful instruction. Competency requires consistent ability to meet technical goals of the intended procedure and to correctly perform the cognitive aspects of the procedure.

The ASGE has published guidelines for training (http://www.asge.org) which emphasize several principles about the process of achieving competency in endoscopy [6]. First, all major procedures require formal preceptorship programs lasting more than several weeks, which are best accomplished in a certified fellowship training program or residency. A minimum number of procedures must be performed, although recent statements acknowledge that the number alone does not guarantee competency. A body of data to be discussed later suggests that the published minimum numbers are too low to achieve acceptable success rates in most instances [7–9]. The endoscopist must be able to achieve a standard rate of technical success, but must also notice any abnormal pathology, identify it correctly, and decide upon the appropriate course of action. Specific training in the administration of conscious sedation and the management of patients receiving such medication is also required.

Linking diagnosis and therapy   

Increasingly, standard definitions of competency for diagnostic procedures demand proficiency in any related therapeutic maneuver that might be required during a diagnostic examination. For example, one cannot be considered competent to perform colonoscopy unless one can consistently reach the cecum, recognize and properly identify a polyp, and successfully remove the polyp during the examination. Next, competency in one procedure does not imply competency in another. Nor does competency in elective procedures constitute competency to perform an emergency therapeutic procedure.

How competent?   

How successful must a trainee be before qualifying as competent to perform a procedure independently? Data on procedure success rates by experts provide a useful perspective in setting the bar. It is seldom possible for a trainee to be as good as a practitioner with several years of experience at the completion of formal training. The ASGE has suggested that trainees be able to demonstrate a minimum procedure success rate of 80–90% by the time they end their fellowship program [10]. The growing body of information detailed below on how many cases are required to reach this level of expertise for each procedure type serves not to obviate the measurement of actual success rates and personal outcomes, but to help ensure that training programs are designed to provide sufficient experience to achieve these goals. Given the objective data available on the learning curve for endoscopic procedures, and given the number of cases performed during the typical 3 year fellowship, an 80–90% success rate appears to be a feasible standard for most procedure types.

Varying rates of learning   

Fellows may learn at different rates and this may depend on case volume, quality of instruction, and on the intrinsic mechanical ability of the particular trainee. There are data to support this principle with regards to acquisition of competency in colonoscopy [O. Cass, personal communication from ACES study data, manuscript in preparation]. The fact that some individuals take more or less time to learn skills than others strongly supports the trend within training programs to emphasize objective criteria for assessing competency rather than relying upon completion of a preset number of total procedures.

Outcomes   

Data are limited proving that individuals who reach objective parameters of competency in training go on to have better patient outcomes than do individuals who begin to practice before such success rates are achieved. Optimal future validation of good training may need to assess not only success rates for intended procedures and complication rates but also patient satisfaction measures.

Learning beyond the training period   

As more objective competency criteria are adopted for trainees in formal training programs, some attention needs to be given to address the issue of practicing endoscopists learning new techniques. They also need objective criteria of competency akin to standards set for fellows in formal training programs for the purpose of credentialing.

 Training and competency in specific endoscopic procedures   

Esophagogastroduodenoscopy (EGD)   

When new GI fellows begin their training, EGD is often the first procedure performed under supervision. During this time they must learn many things besides passing the instrument from the mouth to the duodenum. This is the time they learn about the care and handling of the instrument, the workings of the dials and channels, and the preparation, monitoring, and sedation of the patient, as well as the technical and cognitive aspects of the EGD procedure itself. During the course of a 3 year training program, GI fellows in the USA invariably perform far more than the number of EGDs required to become proficient. However, a number of non-gastroenterologists also perform EGD in the United States. By some estimates, as much as 50% of all endoscopic procedures in this country are performed by non-gastroenterologist health care providers [11]. For this reason, the criteria for proper training in this procedure and evidence to define how much specific training is required to gain sufficient proficiency carry added importance.

Published guidelines for training in EGD   

Most published guidelines for training in EGD are derived from expert opinion in the absence of objective data. Current ASGE recommendations reflecting some recent studies on the learning curve for endoscopic procedures state that at least 130 upper examinations are required to become competent at performing EGD [10]. There is a wide variation in the minimum number suggested by various other professional societies and expert opinions ranging from 25 to 300 cases [12–23]. This degree of variation may result from differences in the definition of what constitutes a competent examination and from lack of outcome data to support the recommendations. A summary of recommendations on minimum numbers of EGD, colonoscopy, and ERCP is shown in Fig. 1.

Defining competence for EGD   

Technical competence in EGD requires the consistent ability to intubate the esophagus with the endoscope and to traverse the pylorus into the duodenum. The entire stomach must be visualized, including a turnaround view of the fundus and cardia. Any abnormal findings must be identified and correctly characterized. Patients must be adequately sedated and appropriately monitored. A trainee must be able to perform targeted mucosal biopsies when indicated.

What technical success rate constitutes an acceptable level of competency? Cass et al. conducted a single-institution study of GI fellows and surgical residents and found that both groups of trainees achieved over 95% pylorus intubation rates after a mean of 120 cases, whereas experienced proctors completed the examination in over 99% of cases [8]. A review of 2500 EGDs performed by family practitioners in the US reported a 93% rate of reaching the small intestine, perhaps reflecting less extensive time devoted to training and considerably lower case volume than gastroenterologists, with an average caseload of 0.6–10.8 cases per month [24]. With readily available access to endoscopists with over 95% success rates, it is difficult to set the bar for achieving competency at technical success rates much lower than this level.

Data on acquisition of competency in diagnostic EGD   

Two studies have addressed the question of the rate of skills acquisition in EGD. In the only published article on this subject, Cass et al. performed a prospective evaluation of seven GI fellows and five fourth-year surgical residents during their training in EGD and colonoscopy. Trainees were graded on standardized forms which tallied objective criteria of procedure success, including both technical and cognitive components. During this study, trainees were graded on a median of 113 EGDs (54–162). Esophageal intubation rates were 95% after 50 cases but then dipped to 75% until completing over 100 cases, reflecting the tendency of proctors to allow trainees to attempt more difficult cases only after increased experience. This finding illustrates a particular feature of objective analyzes of learning curves during training, in which patient mix and case difficulty need to be taken into consideration. For this reason, the authors elected to identify those threshold case volumes after which trainees maintained the success rates designated as minimum standards for competency. Sustained ability to traverse the pylorus was 100% after 90 procedures [8].

In a much larger multicenter study using the same grading form, Cass et al. prospectively studied 135 first-year GI fellows from 14 centres. Preliminary results from this project are published in abstract form, and the manuscript containing the final results is currently in preparation [7; O. Cass, personal communication]. Both objective measures of procedure success and subjective rating of case performance were recorded. For the objective evaluations, competency for a particular procedure was defined as meeting all four of the following criteria: intubating the esophagus, intubating the pylorus, recognizing abnormalities, and correctly identifying any abnormalities. Competency for each trainee in a procedure was arbitrarily defined as ability to achieve 90% success in two successive blocks of 10 procedures. Similar assessments and definitions were used to evaluate the trainees' progress in colonoscopy, as will be discussed below in greater detail.

Data were available for evaluation on 13 195 EGDs. Over 160 EGDs were required before the average fellow met the definition of competency in diagnostic EGD by the objective criteria outlined above. Comparison of data sheets with fellows logs revealed approximately 30% of cases were not scored, suggesting that 160 is an underestimate of the true number of cases required to gain full proficiency in this procedure. Further, those cases which fellows were not allowed to attempt were not included in the evaluations. For EGD evaluation, scoring using the four objective criteria correlated well with subjective ratings of competency on a five-point scale. These data establish that when the components of procedure competency are carefully assessed, the minimum number of supervised training cases far exceeds published guidelines.

These data should not affect GI fellowship programs where fellows routinely perform 500 or more EGDs within the 3 years experience. In contrast, these data directly challenge current practice by non-gastroenterologist providers, who routinely perform far fewer cases during training. One review of experience in surgical training described that surgical residents performed an average of 75 EGDs and 75 colonoscopies during their residency. Family practitioners who perform EGDs may undergo very limited training during short weekend crash courses involving in some instances fewer than 10 supervised procedures [25].

Competency and EGD outcome   

Does additional training make a difference in terms of procedure outcome, complication rates, or patient satisfaction? It is intuitive that individuals with more extensive training and those with higher case volume will have different outcomes than those with less cumulative and ongoing experience. Some data support this notion of variations in outcome. Data from Rodney describing a 93% duodenal intubation rate fall short of the 99% success rates by experienced endoscopists in the Cass study [8,19]. One comparative prospective survey of EGD complications showed that internists had a rate more than three times higher than their GI counterparts [26].

Therapeutic EGD techniques   

(Biopsy, PEG placement, stricture dilation, achalasia dilation, hemostasis techniques, endoscopic mucosal resection, stent deployment, pneumatic dilation, tumor ablation.)

Standard upper GI endoscopy techniques   

The ability to perform targeted endoscopic mucosal biopsy should be considered integral to the performance of diagnostic EGD. Therapeutic procedures such as percutaneous endoscopic gastrostomy (PEG) placement, hemostasis of upper gastrointestinal hemorrhage, and stricture dilation require additional training. One can be privileged in EGD without being capable of performing these interventions. There are not many data available on the amount of training needed to gain proficiency in the therapeutic interventions possible during upper endoscopy. In the absence of data to address the question of how many supervised cases are required to master each of these skills, current recommendations rely on expert opinion. The ASGE has provided guidelines for minimum numbers for some of these procedures [6,10,27,28] (See Fig. 2). Gastroenterology fellows receive ample experience during their training in such procedures as PEG placement, foreign body removal, stricture dilation, and hemostasis of variceal and non-variceal GI bleeding.

Hemostasis techniques   

This topic warrants separate attention for several reasons. First, these cases represent some of the most critical situations facing the endoscopist and can be accompanied by high risk of patient morbidity and mortality. During acute GI bleeding episodes, the endoscopist must combine a number of cognitive and technical maneuvers to successfully diagnose and control the hemorrhage while supervising the management of a potentially unstable patient. Time is a factor and the staffing conditions may be suboptimal, requiring the endoscopist to possess a thorough understanding of equipment set-up and accessory use. These conditions which demand operator proficiency also make the training process more difficult. Cases present at variable rates and times and conditions, which are not always conducive situations for a trainee to spend much time struggling with the endoscope as the mentor patiently provides advice.

Simulation   

Recent work with the compact-EASIE endoscopy simulator (Figs 3, 4, 5) has demonstrated that intensive workshops on this model can lead to significant improvement in trainees' hemostasis skills. This pilot study, presented at the 2002 Digestive Disease Week, characterized the learning curves on the simulator of three specific hemostasis techniques—band ligation, injection and coagulation of bleeding vessels, and hemoclip application—during three weekend workshops over 7 months. In this study, the model was used both as a means of providing repetitive, controlled, realistic practice with expert teaching and as a method of objectively evaluating those skills that were taught. Twenty-eight fellows were randomized to undergo either standard fellowship training or three 7 h workshops in a 7 month period. Baseline skills were compared to final evaluations on the simulator after 7 months. Significant improvements were achieved in all techniques in the simulator group, but only for band ligation in the control group [29]. For the intensive group, the learning curves for the three specific techniques during the project are shown in Figs 6, 7. Evaluations on the simulator after each session revealed a steady improvement in complex tasks such as hemoclip application and injection/coagulation. In contrast, band ligation skills reached maximal levels after only one intensive session. These data provide evidence that certain techniques require more repetitive training than others to master [J. Hochberger, personal communication, manuscript in preparation]. There are no published data yet on the learning curve for technical competency in hemostasis on real patients, or evidence that correlates skill measured on a simulator with successful performance on real bleeding cases.

Bleeding team   

Some institutions have created specialized bleeding teams that direct the hemostasis cases to individuals who have considerable experience in these techniques [30]. Stevens et al. compared the rebleeding rates, need for surgery, hospital length of stay, and hospital charges for patients presenting with acute ulcer bleeding and found that those individuals treated by a specialized 'bleeding team' had significantly improved outcome for each variable as compared to those patients treated by other gastroenterologists at the institution [30].

Retaining competence   

Notwithstanding the potential benefits of subspecialization, this development will certainly raise important question about competency in hemostasis skills in the years to come. Who is competent to perform these procedures and how much ongoing experience is needed to maintain skills [31]? If data are just emerging about the training required to become competent in specific techniques, then data about experience and refresher courses needed to retain competency are even less available. The potential role of simulator-based refresher courses on retraining and skills maintenance will be discussed below.

Other specialized therapeutic upper GI endoscopy techniques   

Opportunities during fellowship to perform other procedures such as pneumatic dilation for achalasia, tumor ablation, endoscopic mucosal resection (EMR), and placement of esophageal stents may vary at different institutions depending on case volume and local expertise. Certain therapeutic techniques need not be mastered by every endoscopist. Nor is every fellow required to become proficient in every technique. For example, some practitioners may elect to refer cases of tumor palliation to individuals with a special interest in this field who are well versed in various endoscopic palliation modalities. Similarly, given the well-recognized risk of esophageal perforation following pneumatic achalasia dilation, it is reasonable for a gastroenterologist with limited experience in this area to opt to send such procedures to a colleague who performs a higher case volume, in the interest of patient safety and improved outcomes. Future data on the learning curves for developing technical and cognitive expertise in these particular specialized procedures will be useful for individuals intending to pursue these interests.

Flexible sigmoidoscopy   

Training in flexible sigmoidoscopy is a topic of particular importance, given the applicability to many non-gastroenterologist practitioners, and even to certain non-physician providers. While a number of other endoscopic procedures are performed by non-gastroenterologists, flexible sigmoidoscopy by such groups is widespread. The future role of sigmoidoscopy in colorectal cancer screening, with the increasing practice of screening colonoscopy and the advent of virtual colonoscopy, is somewhat uncertain. For the time being, given the logistic difficulties inherent in screening the population at risk with full colonoscopy, training in flexible sigmoidoscopy remains an important endeavor. It is unclear what percentage of medicine and family practice trainees currently get trained in this procedure during their residency.

Published guidelines for training in flexible sigmoidoscopy   

The ASGE has recommended that a minimum of 30 supervised examinations be performed in order to develop sufficient proficiency in flexible sigmoidoscopy to perform these procedures independently [6,10].

The seminal paper on the subject of endoscopic training pertains to flexible sigmoidoscopy. In 1986, Hawes et al. conducted a study of 25 residents in internal medicine, surgery, and family practice, in which the trainees were evaluated on consecutive supervised sigmoidoscopies [32]. In addition to grading trainees on a six-point subjective scale, the authors scored the ability of each trainee to visualize each quadrant every 10 cm of the examination, and the performance of the student in correctly making major and minor diagnoses. Eighty percent of residents performing 26 examinations and 89% of those performing over 30 procedures were deemed sufficiently competent by the preceptors to perform independent screening sigmoidoscopies. Several analyzes demonstrated that the individuals with more competent subjective grades performed significantly better on the various objective technical and cognitive parameters measured. For example, polyp detection was 93% once competence was reached. None of the trainees were followed much beyond 30 cases to determine whether any incompetent examinations occurred after a resident had already been determined to be competent by the instructor. There was no requirement in this study for a set number of consecutive competent grades to meet criteria for overall competency. Few of the procedures graded in the competent range achieved the top mark, defined as comparable to a second year gastroenterology fellow, suggesting that the learning curve does not stop after 30 procedures [32]. The numbers reported in this study serve to define a minimum acceptable supervised case experience for trainees in flexible sigmoidoscopy. It is interesting that among teaching institutions in the US at which non-physicians perform flexible sigmoidoscopy, the individuals are often required to perform far more than the 30 supervised examinations recommended by the ASGE, with the number ranging from 25 to 150 procedures [33].

The Hawes study described the learning curve for flexible sigmoidoscopy. It also provided a template for subsequent investigation to study the acquisition of endoscopy skills during training using objective assessment of performance based on objective technical and cognitive parameters of procedure success.

Colonoscopy   

Increased demand for screening colonoscopy warrants a greater focus on the training required to gain proficiency in this procedure. To the extent that better-trained operators perform more complete examinations, cause less patient discomfort, miss fewer important lesions, and cause fewer complications than less well-trained individuals, colonoscopy training may have an important direct impact on patients. Further, these outcome variables influence patient behavior and may affect their compliance with screening recommendations. Inasmuch as debate about potential cancer screening strategies centres around these outcome parameters, training issues in colonoscopy take on added importance.

Published guidelines for training in colonoscopy   

There has been considerable discrepancy in the number of supervised cases recommended by different societies (Fig. 1). On the low end, internist and internal medicine training directors estimated that as few as 25 cases were required to gain proficiency [13,14]. Several gastroenterological societies, including the ASGE, the British Society of Gastroenterology, the European Diploma of Gastroenterology, and the Conjoint Committee for Recognition of Training in Gastrointestinal Endoscopy of Australia, have recommended that at least 100 supervised colonoscopies be performed during training [10, 20–22]. The surgical society SAGES has suggested a minimum number of 50 procedures [18]. Most of these figures were derived from the opinion of practitioners and training directors prior to available data to support such guidelines. In fact, a body of evidence, which will be discussed below, now indicates that all of these minimum training numbers are gross underestimates of the amount of training required before trainees are able to perform colonoscopy at minimum acceptable success rates of 80–90%. More recent ASGE statements reflect a growing consensus that competency in any endoscopic procedure requires the demonstration by trainees of actual proficiency in that procedure, and that no preset number of supervised procedures guarantees that an individual is adequately trained to perform the procedure independently [6,28].

Defining competence for colonoscopy   

Traditionally, endoscopy instructors have considered reaching the cecum without assistance to be the main indicator that a trainee has performed a competent examination. As noted above, the ASGE guidelines suggest a standard of 80–90% technical success rates for individuals completing formal training. Expert gastroenterologists reach the cecum in over 95% of examinations. In a prospective VA multicenter study, gastroenterologists were successful in reaching the cecum in 97.7% of 17 732 colonoscopies [34]. Similar results were reported in a prospective German study [26]. Probably the clearest indicator of competency is the ability to correctly identify polyps and cancer. Expert gastroenterologists should expect to detect colon cancer > 97% of the time [35]. Besides the ability to correctly identify pathology, the proper recognition of cecal landmarks is essential to prevent the false impression of a complete examination.

Technical components   

Technical components of standard colonoscopy include the following:

• Passing the splenic flexure.
  
• Intubating the cecum.
  
• Visualizing 360° of the lumen on endoscope withdrawal.
  
• Providing safe conscious sedation and monitoring.
  
• Maintaining patient comfort during the examination.
  
• Performing targeted biopsies when indicated.
  
• Removing and retrieving polyps when present.
  

Cognitive objectives   

Cognitive objectives essential for competency in colonoscopy include the following:

• Recognizing an abnormality is present or that a normal examination is normal.
  
• Correctly identifying abnormal pathology.
  
• Correctly assessing proper management strategies for the abnormality.
  
• Recognizing and managing immediate complications, including perforation.
  

Minimum training requirements to achieve competency for colonoscopy   

How much training is required to be able to consistently meet these specific objectives? The simple answer is, far more actual supervised cases than have appeared in the listed minimum required numbers in the various guidelines discussed above. The published data are summarized in Fig. 8[7,8,36–40].

The Cass study   

The most comprehensive study addressing this question is the multicenter ACES study performed by Cass et al. [7]. In this study, 135 first-year gastroenterology fellows from 14 training programs were evaluated prospectively on 8349 colonoscopies by 243 faculty preceptors. Subjective assessment of competency on a five-point scale and objective measures of procedure success were recorded for each colonoscopy on standardized data forms. The objective parameters assessed included not only ability to pass the colonoscope to the cecum but ability to recognize and correctly identify abnormal pathology during the examination. Learning curves were constructed. Technical success rates at reaching the cecum of at least 90% did not occur until a median of 195 procedures. Comparison with fellows' logbooks indicated that up to 30% of procedures were not recorded on data forms, implying that the numbers obtained were underestimates of the true number of colonoscopies required to achieve competency. In contrast to the objective measurements, subjective ratings indicated competent ratings > 90% after a median of only 70 cases. This is the best evidence that objective measurement of procedure outcome in endoscopy cannot be surmised by subjective assessment.

Cass performed a least-squares regression of logarithmic curve using the published data in the literature of cecal intubation rates at increasing level of experience, and arrived at an estimate of 341 colonoscopies for the average trainee to achieve a 90% technical success rate [41].

Conclusion   

Competency in colonoscopy requires cognitive accuracy and ability to perform polypectomy. As long as objective competency levels of 80–90% are the endpoint of training, then previously published minimum supervised case numbers are far too low.

Competency and colonoscopy outcome   

The best indication of successful colonoscopy training is good outcome data for practitioners following their supervised period of instruction. The rate of technically complete examinations and complication rates of recent graduates can be compared to published rates for expert endoscopists in the community. Outcome databases generated from computerized procedure reports will make rates of complete examinations easier to track. Other important outcome parameters such as complication rates and rates of missed pathology will be much harder to follow outside of a prospective study. Finally, patient satisfaction data are needed to look at the influence of greater operator experience on this important influencing factor on future screening compliance.

Acceptable outcomes   

With as many at 700 cases performed during gastroenterology fellowship, the suggested 90% technical success guideline for practitioners at the end of training should be feasible, and lower success rates by any practitioner, regardless of specialty, should no longer be considered acceptable. When practitioners experience substandard colonoscopy outcomes, it suggests that either initial training was insufficient or current case volume is too low to maintain adequate skills.

Non-gastroenterologists   

Data on non-gastroenterologists indicate that adequate outcomes may be achieved only after a large cumulative experience is acquired. One family practitioner achieved only a 54% cecal intubation rate during his first 293 cases [42]. At this many cases most gastroenterology fellows reach the cecum over 90% of the time. It is likely that the vastly different daily procedure volume and the benefit of much greater supervised training accounts for this difference. Other internists and family practitioners have reported complete examination rates ranging from 75 to 91% [26,43–45]. Hopper et al. achieved only a 75% cecal intubation rate during 1048 initial cases without formal training, but over 90% if conscious sedation was used [44]. It is expected that recent trainees take time to go from 90% success rates on graduation to the 97% success rates achieved by experienced endoscopists. However, while 90% is an acceptable competency level for a newly trained endoscopist, it is below community standards for an experienced practitioner. One large German study showed that internists had not only significantly lower cecal intubation rates (91% vs. 97%) but also higher complication rates (1 in 1539 vs. 1 in 5155 cases) than gastroenterologists [26]. One study by Rex et al. demonstrated significant differences in sensitivity in detecting colon cancer between gastroenterologists (97.3%) and other practitioners (87%) [35]. Haseman et al. observed a fivefold increase in missed cases of colorectal carcinoma among non-gastroenterologist colonoscopists as compared to gastroenterologists; of the missed cases, approximately half were due to inability to reach the cecum [46].

Excellent outcomes are not limited to gastroenterologists, however. Wexner presented a procedure success rate of 96.5% for four expert surgeon colonoscopists [47]. In the initial single center study by Cass et al. seven GI fellows and five fourth-year surgical residents had similar performance levels [8]. These data suggest that specialty does not influence objective measures of competency, as long as the training and case experience is comparable and sufficient to achieve acceptable success rates.

Rate of skills acquisition for colonoscopy   

The issue arises that some trainees learn faster than others. As training program directors look for more efficient methods of training to meet increasing staffing needs for colonoscopy, factors which influence the rate of skills acquisition will be important to define. In the Cass study, there were significant differences in the mean number of procedures performed, before becoming competent, between the fastest and slowest 5% of fellows [O. Cass, personal communication, manuscript in preparation]. The baseline hand–eye coordination of the trainee may influence the early part of the learning curve as the fellow learns to control the dials and torque the instrument. If this could be assessed at the start of training, those will less developed skills might benefit from additional exercises on simulators or static models to accelerate their learning rate, or at least make the initial cases more tolerable for the patients.

Cases per week   

The volume of cases performed per week may also influence the pace of skills development. The training environment is often shaped by the clinical activity of the endoscopy unit, but more information about the optimal caseload per week from the trainee's perspective may reveal that some fellows are performing too few and some too many cases each week. Alternatively, given increased workloads in some endoscopy units, it is quite possible that some trainees actually spend too much of their time performing colonoscopy. At this author's institution, data from fellows' logs indicate that the trainees perform close to 700 colonoscopies during their 3 years' experience.

Too many cases?   

Many of the studies on training focus on the important question of how much procedure training is enough. This is most appropriate in a setting where non-gastroenterologists are interested in performing these procedures without completing a dedicated training program akin to that of the gastroenterology fellow. But an equally valid question which has not received much attention or investigation is how much is too much, when more time might be devoted to developing cognitive skills, knowledge base, or research interests.

Therapeutic colonoscopy (biopsy, polypectomy, hemostasis techniques, stricture dilation, stent deployment)   

There is scant literature on the specific amount of training required to gain proficiency in therapeutic techniques performed during colonoscopy. Published guidelines acknowledge that individuals require extra specific training to develop these skills. Past recommendations have specified a suggested minimum number of procedures for trainees to perform [10,27], but there is really no evidence to support that these numbers are sufficient (see Fig. 2). Most of these therapeutic colonoscopy procedures are routinely taught to GI fellows, and training programs typically provide trainees with ample case volume to surpass these numbers during a 3 year fellowship.

Standard therapeutic techniques (integral to performance of diagnostic colonoscopy)   

Given the widespread access to practitioners with the ability to perform targeted biopsy and polypectomy whenever the need arises during a diagnostic endoscopy, some have questioned the ethics of having anyone perform diagnostic colonoscopy without sufficient training to take biopsies or remove polyps [48]. Non-gastroenterologists who provide colonoscopy services should be able to not only achieve cecal intubation rates that are comparable to community gastroenterologists but should also be able to perform these common therapeutic maneuvers to avoid unnecessary repeat procedures. It could also be argued that, for very large polyps, if the endoscopist is not comfortable attempting to remove them endoscopically, then the individual must have the ability to mark the lesion with a tattoo to facilitate subsequent lesion localization by either an expert endoscopist or surgical consultant.

Advanced therapeutic colonoscopy techniques   

Management of large polyps, saline-assisted polypectomy, endoscopic dilation of colonic strictures, and placement of self-expanding metal stents for tumor palliation are procedures requiring specific training and expertise. Unlike the ability to remove polyps that are encountered during the examination, it is not imperative that every practitioner performing colonoscopy be capable of performing these advanced techniques. GI fellows typically receive some training in all of these procedures, although the extent of their experience during fellowship may vary. No data exist defining the amount of supervised training needed to become proficient in these specialized areas.

Diagnostic and therapeutic ERCP   

Published guidelines for training in ERCP   

The ASGE has recognized that certain endoscopic procedures, such as diagnostic and therapeutic ERCP, are more complex and entail greater risk than standard procedures. Accordingly, in 1994 this organization developed guidelines for advanced endoscopy training, and updated them in 2001 [49]. Current minimum numbers of ERCPs which must be performed before competency can be assessed include 200 cases, including at least 40 sphincterotomies and 10 stent placements [10]. Notable among the general principles set forth in these recommendations are the following:

• Those training programs which provide advanced endoscopic training must have sufficient case volume and local expertise.
  
• The former model of providing trainees who have 100 or fewer procedures with subsequent additional supervised training during practice is no longer viewed as acceptable.
  
• Training should be focused upon trainees who will get to perform a large volume of cases, with the expectation that they might develop sufficient skill to perform ERCP independently at the completion of the training period.
  
• Objective monitoring of performance should occur during training [49].
  

Not all programs are suited to offer such training and not all fellows should receive this training.

Non-technical training   

The published training guidelines emphasize the non-technical aspects of ERCP training and recommend that substantial time is devoted to this pursuit. Time spent attending meetings and participating in endoscopic research is also advocated. While it is possible for someone with sufficient intrinsic ability and caseload to complete this training within a 3 year GI fellowship program, many trainees will require a dedicated fourth year of training to gain technical and cognitive proficiency in ERCP.

Defining competence for ERCP   

ERCP involves a wide array of different diagnostic and therapeutic techniques. To be competent at ERCP, a trainee must be able to consistently perform the intended procedures without assistance. The ASGE guidelines suggest that trainees be able to perform all of the following tasks [6]:

• Cannulation of desired duct.
  
• Opacification of desired duct.
  
• Stent placement.
  
• Sphincterotomy.
  
• Stone extraction.
  

Technical success   

ASGE recommendations are that trainees be able to perform these techniques with >= 80% success. Similar to other procedures, this must be accompanied by concomitant demonstration of appropriate interpretive and diagnostic proficiency.

As with colonoscopy, the appropriateness of outcome-based definitions of competency in ERCP can be validated by examining the success rates of expert practitioners. Expert endoscopists have been shown to successfully cannulate the common bile date at a rate of > 95% [50–53]. One study from England described a range of biliary cannulation success rates among practitioners of from 76% to 95% [54]. It is problematic to generalize results published from large academic centers, which are subject to a referral bias in which more difficult cases are seen at expert centers. Such data may also be skewed in the opposite direction due to operator bias in which more experienced endoscopists at high-volume centres are more willing to utilize aggressive techniques in order to achieve successful intervention than are less experienced colleagues in the community. All information on ERCP outcome that looks at both procedure success and complication rates must consider these influential forces.

Limited information is available for standard benchmarks for specific technical aspects of ERCP, apart from achieving biliary access. One can extrapolate, from published series in the literature, the specific procedure success rates at expert centers for sphincter of Oddi manometry, mechanical lithotripsy, deployment of metal biliary stents, pancreatic endotherapy, and minor papilla endotherapy. However, less information is available about success rates for these techniques by endoscopists in the community.

Varying case difficulty   

Measures of technical success, such as selective cannulation, can be recorded, but how does one rate a trainee when the expert proctor has difficulty with or fails to achieve the procedure objective? Schutz et al. have devised a rating scale to describe the degree of difficulty intrinsic to the procedure [55]. Both the nature of the intended procedure and the degree of difficulty actually experienced by the tutor when the trainee is unable to complete the case provide relevant information about the case difficulty which should impact on any rating of trainee performance. Future efforts to prospectively evaluate trainees' progress during training or to objectively assess their competency level at the end of training ought to take these factors into account.

Judgement   

The highly subjective assessment of endoscopic judgement also influences the overall competency rating of a fellow at the end of training. Given the risk of severe complications, it is sometimes the best medical decision for even an expert not to persist in an attempted difficult procedure if the indication is limited. For example, failure after a short period of time to achieve deep biliary cannulation in a young female patient with non-dilated ducts and suspected type I sphincter of Oddi dysfunction generally does not warrant more vigorous attempts to obtain access, and certainly not precut techniques. Evaluation of competency in ERCP must somehow manage to take judgement into account in addition to technical success, cognitive skills, and procedure difficulty.

Minimum training requirements to achieve competency for ERCP   

Case numbers   

Although previous guidelines had established a minimum threshold of 100 diagnostic ERCPs to achieve competency [56], one study by Jowell et al. demonstrated that 180 such cases are required to cannulate the desired duct with a success rate of at least 80% [9]. In this study, proctors prospectively graded 1796 ERCPs performed by 17 fellows. Overall ERCP competence was defined as having a rating of 1 or 2 on a 6-point subjective scale for 20 consecutive procedures. The authors found that the probability of achieving an acceptable score of 1 or 2 out of 6 for technical components without assistance did not surpass 80% until 185 ERCPs. In this study, proctors were instructed to take procedure difficulty into account when assigning the subjective score. The number required to achieve deep biliary cannulation and sphincterotomy with competence did not reach 80% at 180 cases [9].

One limitation of this study is the fact that only three fellows performed over 180 ERCPs during the study period. Therefore, while it is difficult to infer that fellows will achieve competency at ERCP by 180 cases, these data demonstrate that previous recommendations of 100 cases grossly underestimated the experience required to master this procedure. Another study analyzed the learning curve of 21 trainees during their first 100 ERCPs and similarly found that no fellow reached sustained acceptable biliary cannulation success rates during the study [57]. Because so few data have been collected on trainees' performance as they reach higher procedure numbers, the true number of ERCPs required to achieve competency remains uncertain, except that it is far greater than 100. Recently revised ASGE credentialing guidelines suggesting a minimum experience during training of 200 cases reflect these data [10]. More important than the revised minimum number of cases required during training is the increased focus on objective outcome of trainees' performance.

What is a case?   

Especially during ERCP, the length of time that a fellow gets to attempt each portion of the procedure without assistance by the preceptor may vary considerably with the clinical situation and the comfort level of the preceptor with the progress of the case. Accordingly, the number of cases that a trainee lists as having 'performed' does not always tell the whole story of his or her degree of involvement in each case. Perhaps more relevant than the standard fellow's case log would be a record of cases in which the trainee cannulated the desired duct or performed the indicated therapeutic procedure without manual assistance from the instructor.

Competency and ERCP outcome   

It is unclear what happens if one begins to practice with less than 80% success rate at the onset in terms of adverse patient outcomes and in terms of eventual development of appropriate levels of competency. More data are needed to correlate data on assessments of trainees' competence with patient outcomes as they begin to perform ERCP and other procedures on their own.

Improving after training   

Schlup et al. described the ERCP learning curve of one endoscopist. In this study, an innovative CUSUM method was used to graphically illustrate the gradual skill improvement for this individual during 532 cases over 8 years, after 2 years of assisting another practitioner. The rate of selective cannulation increased from 85% during the first 2 years of practice to 88%, 90%, and 96% over successive 2 year blocks. A downward slope of the CUSUM curve for 95% successful cannulation rate, reflecting the lack of further incremental improvement in performance with additional case experience, was achieved after 350 cases [58]. While it is hard to generalize the results of one endoscopist, this study demonstrated that competence continues to increase with practice following formal training in a fairly steady manner until far more than the recommended minimum number of cases have been performed. Davis et al. looked at cumulative ERCP experience and found that endoscopists with > 200 total prior cases had a higher success rate at sphincterotomy 97–99% vs. 88% [59].

Annual volume   

Indicators of competency such as procedure success rates may depend on annual procedure case volume as much as total prior experience. In 12 low volume centres in which less than one sphincterotomy is performed per week, Freeman et al. reported a failure to achieve biliary access and drainage after sphincterotomy of 5.4% compared to 1.2% for 5 high volume centres [60]. For all ERCPs, endoscopists performing more than two cases per week had significantly higher cannulation rates (96.5% vs. 91.5%) [50]. Loperfido et al. similarly found a failed cannulation rate of 7% vs. 3.9% at centres with low volume (< 200 cases/year) vs. high volume of ERCPs performed [61].

As Freeman has emphasized in several publications, failed procedures result in increased health care resource utilization with its associated costs and morbidity [62,63]. In a review of 10 000 sphincterotomy results at a single center over 15 years, while complication rates did not significantly differ between the first 10 and last 5 years, the need for second attempts to complete the procedure did decrease from 8% to 4% [64].

Competence affects complication rates   

Competency affects not only procedure success rates but also the frequency of adverse events. For this reason, complication rates serve as another marker of competency in ERCP. A number of studies have looked at high volume vs. low volume centres in terms of adverse events, although few data are available comparing endoscopists with more and less overall experience. In a seminal study, Freeman et al. found that high volume centres, defined as those performing >= 1 sphincterotomy per week, had fewer overall complications (8.4% vs. 11.1%), fewer severe complications (1.1% vs. 2.9%), and fewer hemorrhages (1.1% vs. 2.9%), but similar rates of pancreatitis [60]. Other authors have found higher overall complication rates at lower volume centres [50,61,65], or among less experienced endoscopists [66]. Rabenstein et al. found that higher procedure volume (> 40 sphincterotomies per year) was a more important predictor of lower adverse events than was overall case volume associated with years of experience [65].

Rate of acquisition of ERCP skills   

As with colonoscopy, different fellows progress at different rates in skills acquisition. This was confirmed in the study by Jowell, with two fellows reaching competency after 120 ERCPs [9]. Some of this may relate to the innate abilities of the particular trainee. One extrinsic factor which may factor in to the learning curve is the frequency of hands-on work. It remains to be established whether fellows at high volume centres achieve competency at fewer numbers of procedures than their counterparts at less busy institutions. There are data described above to indicate that greater frequency of performing sphincterotomy leads to better success rates and lower complication rates [60,65]. However, no study has indicated yet that trainees learn faster or better when subject to higher daily endoscopy caseloads.

Therapeutic ERCP   

Current ASGE guidelines suggest that a minimum of 40 sphincterotomies and 10 stents be performed before competency can be assessed [10]. There are no published recommendations about other procedures such as bile duct stone removal, stricture management, and pancreatic endotherapy. Eventually, the guidelines may evolve to incorporate more specific recommendations about the expected success rates which trainees should achieve at particular therapeutic techniques before independent practice, and the number of such cases typically needed to achieve these standards.

It is also recommended that 50% of procedures performed during training involve therapeutic maneuvers. As the proportion of therapeutic ERCPs at teaching centres increases, this goal should not be difficult to achieve. The principle of training endoscopists who can perform necessary therapeutic maneuvers at the same setting as the diagnostic procedure is critical. This is akin to the principle described above which considers the ability to perform snare polypectomy to be a required skill for all individuals wishing to perform colonoscopy. This diverges from former community practice, in which some practitioners would attempt diagnostic ERCP with the foreknowledge that any need for therapeutic intervention would require a referral for a second procedure. With widely available expertise in therapeutic ERCP, and given the added risks and costs of delaying therapy and subjecting patients to an additional procedure, this practice is no longer justifiable. If a trainee must be competent to perform sphincterotomy if he is to undertake any ERCP, then even 180 supervised cases may be insufficient, as sphincterotomy success rates >= 80% may not be achieved with that amount of training [9].

This emphasis on exposure to therapeutic techniques may affect the style of advanced endoscopy teaching at some institutions. At some programs, the trainee is allotted a certain amount of time with the endoscope, or allowed to handle the duodenoscope only until reaching an impasse. In this way, opportunities to gain experience with a procedure later in a case, such as stent placement, might be wasted unless the endoscope is returned to the trainee after the tutor performs the cannulation and sphincterotomy.

Rate of acquisition of therapeutic skills   

What are the data on the rate of skills acquisition for the most common therapeutic ERCP techniques? One study looked at the learning curve of sphincterotomy as well as the outcome of sphincterotomy performed by individuals with different levels of experience [65]. In this retrospective analysis of 1335 sphincterotomies, a small subset of three endoscopists had data collected on the outcome of all sphincterotomies performed, starting with the initial case of each and continuing until he or she had performed over 100. Complication rates dropped to below average after 40 cases for this group, leading the authors to suggest 40 as the minimum number to perform during training. While fellows in the Jowell study did not achieve competency in sphincterotomy by 180 cases, the authors did not indicate the mean number of sphincterotomy attempts performed by the fellows during those 180 cases, making it impossible to extract a learning curve for sphincterotomy from this paper [9].

In the Jowell study, in which fellows were allowed to place stents even if the proctor had to perform the cannulation, competency in stent insertion occurred after 60 ERCPs. Competency in stone removal occurred after 120 procedures [9].

More data are needed to support the ASGE minimum recommendations of 40 sphincterotomies and 10 stents, as well as to characterize the learning curves for other ERCP techniques such as stricture management, sphincter of Oddi manometry, and pancreatic endotherapy.

In one recent study[67] Harewood et al. described their experience in 175 patients requiring drainage of pancreatic fluid collections, and demonstrated significantly better success rates between the first 20 procedures and the subsequent cases (45% vs. 93%; p= 0.0002) and a reduction in days to resolution (50 days, initial 20 procedures vs. 33.5 days, subsequent procedures; p= 0.05). Such data both characterize a learning curve and help set the bar for success rates possible by experts in this particular procedure. The only problem with relying on this type of study to generalize benchmarks and training requirements is the probability that other individuals with less prior experience in therapeutic endoscopy might have difficulty achieving such high success rates and might require a much greater experience to maximize their own outcomes.

Diagnostic and therapeutic EUS   

The ASGE has recently published guidelines for training in endoscopic ultrasound [68,69]. Development of proficiency in this procedure requires a great deal of cognitive training to identify pathology accurately. In addition to performing EUS examinations under the supervision of an expert, fellows may need to utilize a variety of didactic tools in order to develop interpretive skills and understanding of the anatomy.

Defining competency in EUS   

Pooled data on reported accuracy for EUS in staging of various tumors may be used as outcome benchmarks for practitioners wishing to perform EUS [58](See Fig. 9). As diagnostic accuracy is the best measure of procedure success rate, efforts to assess competency must include attempts to correlate findings with other staging data, including other imaging modalities, pathology, and operative findings when possible. When this correlation is not possible, then direct comparison with the findings of an expert instructor must serve as a surrogate measure of proficiency. The problem of interobserver variability in interpretation [70] underscores the importance of clinical–pathological correlation whenever possible for both the learning process and the assessment of competency.

Learning curve for EUS   

How long does it take a trainee to become competent to perform diagnostic EUS? A study by Hoffman et al. found that accurate evaluation of the esophageal and gastric mucosa occurred in a median of 10–15 cases. The celiac axis could be evaluated after a median of 25 cases [71]. A survey of the American Endosonography Club suggested somewhat higher numbers of procedures required to accurately evaluate the upper gastrointestinal mucosa [72].

One study of T-staging for esophageal tumors showed that subjects required 100 procedures before being able to achieve acceptable standard rates of accuracy [73]. Another study found an 89.5% accuracy after 75 cases [74]. Expert opinion-based ASGE recommendations suggest that novices can successfully evaluate submucosal lesions after 40–50 cases [6]. Most EUS practitioners agree that trainees must perform considerably more cases to master the more complex anatomical relationships observed in examining the pancreas, with numbers ranging from 121 to 150 cases [72,75]. One abstract, however, found that fellows could correctly identify pancreatic abnormalities after a mean of 54 procedures performed (range 13–134) [71]. Still, there are few prospective data to characterize the learning curve for pancreatic examination with EUS. Published ASGE guidelines on credentialing and privileging for EUS recommend a minimum number of 100 supervised cases for individuals hoping to perform non-pancreatico-biliary EUS, and 150 cases for those wanting to perform all procedures (including 50 fine needle aspiration (FNA) procedures and 75 pancreatico-biliary cases) [69](see Fig. 10).

Therapeutic EUS   

Increasing therapeutic capabilities such as FNA, celiac plexus block, and pseudocyst drainage require specific additional training. A brief period of formal expert instruction may dramatically affect the learning curve for therapeutic EUS. Harewood et al. demonstrated that three experienced endosonographers (> 300 total cases with > 100 pancreas cases) had a significant improvement in EUS–FNA accuracy from 33% to 91% between their first and second 8 months of FNA experience after performing 2–3 cases with an expert mentor at the end of the first 8 month block. While the total number of cases evaluated retrospectively was small (n = 20), the cellularity of specimens was increased and the mean number of needle passes required to obtain tissue was reduced in the latter cases [76]. In a similar study [77], Mertz demonstrated a steady improvement in sensitivity for pancreatic cancer diagnosis with operator experience from 50% during the first 10 cases to 80% after 30 cases. This sustained sensitivity >= 80% from cases 30–80 supports the ASGE recommendation that individuals have at least 25 supervised FNA cases during training.

It is unknown whether high volume experience will prove as efficient as this limited expert hands-on mentoring in achieving this level of proficiency in FNA. More data are also required to define the learning curves for other therapeutic EUS applications. Published guidelines stress that (a) these procedures should not be undertaken until a trainee is fully competent in diagnostic EUS, and (b) that considerable extra dedicated training is required [68,69].

EUS training opportunities   

Opportunities for formal advanced training in EUS are limited in the United States. Programs which offer such training are listed on the World Wide Web at:

http://www.asge.org/gui/resources/manual/eus_training_participants.asp

Twenty-eight academic institutions are listed, 21 of which offer fourth-year dedicated EUS fellowships.

While many gastroenterologists are attending short courses and attempting to teach themselves diagnostic EUS, a satisfactory curriculum and standard means for practicing gastroenterologists to train in EUS has not been established. If the need for endosonographers outstrips the availability of individuals who have obtained sufficient specific training in EUS as fellows, then alternative pathways for skills acquisition will need to be more clearly developed and expanded.

 Complementary methods for instructions in GI endoscopy   

Advances in didactic methods   

Self-instruction   

A variety of didactic materials may be used to complement the supervised mentoring of endoscopic procedures. Standard textbooks and atlases are useful for beginners. Procedure technique videotapes, such as the collection sponsored by the ASGE and available to ASGE members on the Web may have relevance to the experienced endoscopist as well as the beginner, depending on the content of the video. Highly edited videotape offers concentrated attention to key teaching points narrated by expert endoscopists, although the editing can make the conditions appear less realistic than raw live footage. Such items may be incorporated formally into an introduction to endoscopy curriculum at a particular institution, or used independently as references.

A number of new technologies offer expanded possibilities for self-tutorial. A number of CD-ROM, interactive DVD, and Internet offerings are either currently available or under development. For example, several 'introduction to' CD-ROMs are available to help beginners orient to the anatomy as they learn to perform EUS. Electronic textbooks like this one, which take advantage of hyperlink capabilities and utilize the Web for easy access, will allow trainees to explore as their own level of expertise and interests dictate. Self-assessment is possible through this learning modality as well. Other Web-based didactic activities conducive to self-instruction include cases of the month, chatrooms, and Web-casts of live procedures. The developers of some of the computer simulator models discussed below hope to capitalize on Web-based mentor feedback of self-instruction done on the actual simulators at remote sites. The actual importance of 'e-learning' activities in the future of endoscopy training and maintenance of skill remains to be seen, but the possibilities are considerable.

Group instruction   

Review courses for gastrointestinal endoscopy come in many formats and sizes. Lectures replete with still and video illustrations are the mainstay of these offerings. The syllabus may cover a wide array of endoscopic issues and techniques or focus instead on a particular topic. They include large annual postgraduate courses serving 1000 or more registrants, smaller regional courses for a few hundred, and small workshops sponsored by local institutions for a much more limited audience. Unlike the self-instruction modalities, with the exception of Web-based interactive exchanges, attendees have some opportunity to ask questions from instructors. Even so, depending on the size of the gathering, such opportunities may still be limited. Clearly, the small workshop or breakout session of larger courses featuring live commentary of edited videotape is better suited to enable interaction and active learning.

In one study by Laine et al. exposure to a brief didactic teaching period by attendees at a postgraduate course significantly improved their correct characterization of stigmata of ulcer hemorrhage from 72% to 82% overall, particularly among fellows, who had a 15% increase in correct answers [78]. Despite a dearth of similar data demonstrating direct benefit of short didactic presentations, this modality of instruction clearly has an important role as a complement to supervised, endoscopic hands-on experience.

Laboratory demonstrations   

Live endoscopy demonstrations are extremely popular with audiences. They offer a complete view of the interactions in the endoscopy unit and expose the viewers to a wide variety of situations, some unexpected. Seeing a case in real time, hearing how experts think through a problem, and watching how they get out of trouble is invaluable, and not possible on most edited videotape.

Despite the excitement and educational value of this format, a number of disadvantages and limitations of live courses exist [79]. While many patients are happy to undergo procedures by an expert endoscopist whom they have never met, less operator knowledge of individual patients leads to increased risk and ethical concerns. Unless live demonstrations are conducted for a small group of individuals with direct verbal access to the endoscopist, they offer limited interactive opportunities.

Wide variability in the experience level of the attendees makes it difficult to tailor the cases and the discussion appropriately to the learning objectives of the audience. In fact, many of the advanced techniques demonstrated may have limited relevance to the average endoscopist, who may learn more about when to refer a patient to an expert than how he or she should perform a technique. When visiting faculty perform live procedures, there are inherent risks due to unfamiliarity with the particular unit and staff.

There are also potential problems stemming from pressures for faculty to showboat talents and succeed in the procedures at all costs, even when it might be better from the patient perspective to stop. The variability in pathology may present logistical difficulties, but this problem can usually be overcome at high volume centres. Down time is another potential issue, which can be minimized if more than one room can be used simultaneously for transmission. Still, the realistic complete view offered at live courses may come at the expense of efficiency of instruction. These courses are both expensive and labor-intensive to organize and run.

Given the popularity of live courses, a few of these will likely continue. Local smaller live workshops performed by local endoscopists with expert visiting commentators overcome many of the above-mentioned concerns. Such offerings are also likely to prosper in the future but with limited accessibility due to audience size [79–81].

Endoscopy simulators   

Static models   

The concept of using static models to teach basic hand–eye coordination, use of endoscope dials, and even basic pathology recognition, is well established. Schindler described using a model stomach for practice in orientation [82]. A number of static models were developed in the 1970s, including the Heinkel hemispheric anatomical model [83], Classen's upper GI plastic dummy [84], and Christopher Williams' St Marks colonoscopy plastic tubing simulator [85]. More recently, Leung and Chung described use of a static model to teach ERCP [86]. The main drawback of all these models has been their limited ability to recreate realistic conditions. For the most part, these learning devices, designed to introduce trainees to the mechanical manipulation of the endoscope within the lumen, offer limited exposure to varying pathology.

Courses with static models   

In perhaps the most comprehensive use of static models, Lucero et al. conducted a series of courses in which they combined a static model with superimposed painted pictures to represent various commonly encountered abnormalities. These courses combined didactic lessons, slides, videotape, supervised hands-on training on models, a progressive lesson plan with increasingly challenging manual and cognitive tasks, and objective skills assessments of trainees by mentors [87]. The principle psychomotor training program designed by this group is called the SimPrac-EDF y VEE (Simulator for the Practice of Fiberoptic Digestive Endoscopy and Electronic Video Endoscopy). Several types of models apart from the one developed by these authors were employed. For example a specific Bilroth II model was created to introduce trainees to this anatomy.

The Classen and Heinkel models were also used to teach specific skills. The authors delineated a series of learning modules and teaching objectives, and created a lesson plan employing diverse modalities to reach those objectives. During these courses, trainees had a mean duration of hands-on practice of 28 h, with 8–25 individuals included in each particular workshop. Unfortunately, the published report of this wide experience in over 22 such courses involving 422 trainees fails to delineate many details of the observed benefit of such training. The report indicates merely that 95% of trainees demonstrated an acceptable level of skills by the end of the training [87].

Still, this concept of providing intensive multimodality workshops with expert instructors, a preset lesson plan, integration of manual training with cognitive skills, and immediate feedback and evaluation, serves as a model for future efforts employing the more realistic and sophisticated simulators which have since been developed. Each of these new simulators has features which make it most useful in teaching a particular set of skills and best suited for a particular stage of the learning process. Some suggestions for how best to integrate simulators into the typical fellowship program are included in Fig. 11.

Lucero's courses appear to be hugely labor-intensive and difficult logistically to conduct. Discovering which simulators are the most effective means of instruction for different techniques is a current area of active investigation. However, as future efforts are made to improve endoscopy training using simulators, innovators must bear in mind that any revamping of endoscopy education will require that issues of feasibility, manpower, and cost need to be addressed.

Ex vivo artificial tissue models: the 'phantom' Tübingen models   

The next generation of sophisticated static models was developed by Grund et al. at the University of Tübingen in Germany [88]. Rather than integrated painted images to represent pathology, this 'Interphant model', or 'Phantom', employs artificial electrically conductive tissue called Artitex to craft pathological lesions such as polyps and tumor strictures which are sewn directly into the three-dimensional latex anatomical model (Figs 12, 13). These models still lack the realism of real bowel wall compliance and motility.

However, not only is the integrated pathology realistic in appearance, it also allows for the application of electrosurgical interventions. The colon model fixed structure employs a semiflexible series of coils similar to prior static colon simulators, which gives the trainee a degree of resistance to endoscope passage to mimic the feel of an actual procedure. To allow for an even wider possibility of techniques which may be performed, Grund et al. have incorporated real animal tissue into the framework of their model. In particular, they craft an ampulla of Vater using a chicken heart, replete with separate pancreatic and biliary orifices, and insert this into their upper endoscopy simulator (Fig. 13).

When used at a training course, several polyp-laden colons and chicken heart papillae can be prepared in advance and quickly inserted into the chassis of the model after the initially prepared material has been depleted. For this reason, many trainees can work on the same simulator in succession during a training course. Of the existing simulators, this one stands out for its realistic polypectomy, argon plasma coagulation simulation, and ERCP simulation. While fluoroscopy is not integrated, the orientation of this synthetic papilla more closely resembles that of humans than the porcine papilla found in the Erlangen models described below. Pancreatic cannulation and endotherapy is also possible, in contrast to the porcine tissue models in which the pancreatic orifice is not readily accessible. Procedures that require submucosal injection are not feasible.

The major drawback to this type of simulator is that currently they are not mass produced and the pathology remains hand-prepared. For this reason, the 'Phantoms' are not widely available and can only be incorporated into training courses at considerable expense, with the Tübingen team in attendance. Efforts to incorporate this simulator, along with the Erlangen models, into wider endoscopy curriculum innovation within Germany is underway.

Ex vivo animal tissue simulators: EASIE and Erlangen models   

In 1996, Hochberger and Neumann developed an innovative simulator, using pig organs obtained from a slaughterhouse and fastened to a plastic platform [89,90]. To this model, Hochberger added a highly realistic simulation of pulsatile arterial bleeding (Fig. 5), created by perforating the stomachs and sewing in real arteries attached to a roller pump capable of pulsatile perfusion with cherry-colored saline solution. Subsequently, other pathologies have been easily recreated using this model, including polyps, varices, and strictures [91,92].

Two models based on these principles now exist. In the original Erlangen model, the pig organs are inserted into a dummy mannequin. This model may also be used to simulate various laparoscopic surgical procedures [93]. A smaller, portable, lightweight version, the compact-EASIE retains the basic tabletop platform to which only the organs needed for endoscopy simulation are pinned (Figs 3, 4). This may include just the esophagus to duodenal bulb, or may accommodate the liver and hepatobiliary tree to allow ERCP simulation involving fluoroscopy. On either of these animal tissue-based simulators, numerous therapeutic procedures may be demonstrated, taught, and evaluated. Basic skills instruction and hand–eye coordination practice is possible, though this may not be the most efficient use of this versatile model.

Perhaps the most well known application of these simulators is in hemostasis training. The results of the pilot validation study of the efficacy of EASIE workshops was discussed previously. A larger European study is currently underway. The Hochberger group has conducted a number of training courses using the EASIE model in other areas, including endoscopic mucosal resection (EMR), stricture management, vital staining, polypectomy, and ERCP [92,94–96]. The techniques can vary from basic biliary cannulation and plastic stent insertion to choledochoscopy, laser lithotripsy, and placement of bilateral hilar metal stents.

The set-up and preparation of workshops using this technology is extensive. While the compact-EASIE promises to facilitate dissemination of simulator training to local sites, widespread training of trainers to prepare the stomachs and run independent training workshops has yet to occur. Similar to the experience of Lucero et al. described above, training sessions using the Erlangen or EASIE models are labor intensive, with high faculty to trainee ratios (Fig. 14).

One key advantage of this model is the facility for trainees to perform multiple repetitions of the same technique. The use of real tissue and the capability of performing advanced therapeutic procedures make this the current simulator of choice for practitioners hoping to learn new techniques, and for second- and third-year fellows hoping to get some non-pressured experience performing interventions to complement their unpredictable supervised exposure to real therapeutic endoscopy. Limited availability and logistic concerns remain obstacles to more widespread use of this important new method of endoscopy training.

Live animals   

Endoscopy has been performed on live anesthetized pigs and dogs for both research and training courses [97–99]. Such experience provides the best possible tactile 'feel' of real tissue and endoscope movements with conditions most closely resembling real endoscopy. This includes the presence of luminal fluid and motility, and the ability to cause real bleeding and perforation. Live animal courses have been conducted to teach therapeutic techniques, most notably ERCP and EUS [100]. Using live animals is the only non-human means of simulating sphincter of Oddi manometry [101].

Despite these advantages, there are several reasons why such live animal courses are likely to have only a limited role in the future of endoscopy training. Animals are very expensive to maintain and there are serious ethical considerations in using them for training, especially when alternatives exist for teaching most techniques which do not require sacrificing any animals for this purpose. Such courses can only be held in specialized animal facilities. Only a limited number of attendees can participate in any given workshop. Once certain procedures such as sphincterotomy have been performed, it is difficult for others to practice the same techniques on the same animal. The high costs require registration fees of over $1000 per session. For these reasons, training on live animals, while potentially more realistic than inanimate simulators, is neither an efficient nor a cost-effective means of education.

When such courses are held, it is important that the skills taught are limited to those techniques not well taught on inanimate models and that attendees be at the appropriate level of training to derive the maximum benefit from the experience. It is not justifiable to sacrifice animals to provide novices with an opportunity to try out advanced techniques which they are unlikely to ever perform in their clinical practice.

The porcine anatomy leads to a few particular considerations which affect the training. First, the long stomachs often require 2 days of fasting to clear, and poor visualization due to undigested food can be problematic. The biliary papilla is located only a few centimeters beyond the pylorus, making it difficult to achieve cannulation in the standard 'short scope' position. The pancreatic papilla is located separately distal in the duodenum, making pancreatic work difficult in the live pig model. One advantage to the porcine anatomy is the presence of a firm polypoid protrusion at the pylorus, the 'torus pylorus', which provides a nice target for practice using a needle knife [92].

In the future, live animal courses will be limited to advanced procedures such as EUS and manometry for which no comparable inanimate models exist. Live animal endoscopy labs remain well suited for clinical investigation. Testing of new accessories and development of new techniques on live animals will continue, but a lot of the early work will be more efficiently carried out on inanimate simulators.

Computer simulation   

The field of computer simulation of endoscopic procedures is rapidly developing. A number of investigators have pioneered efforts to produce models which could allow realistic experience handling the endoscope and incorporate a broad exposure to pathological images [85,102–111]. In theory, this modality has several advantages for the trainee over some of the other educational methods discussed above. As numerous educators have recognized, computer simulation offers the best opportunity to expose trainees to a wide range of pathology. After the initial purchase cost, using computer simulators requires no set-up, and learning can take place either independently or as part of larger training courses. Progressive tutorials of increasing difficulty can be constructed. Progress during training can be recorded and opportunities for feedback exist. Unlimited repetition and drilling in specific infrequently encountered procedures is possible.

Like the static models, most computer simulators utilize an endoscope passed into a dummy mannequin. Recent advances in tactile feedback capability have greatly enhanced the realism of endoscopy simulators. Sensors on the endoscope tip generate the sense of tactile feedback. The incorporation of real video images serves to further enhance the experience. It is possible to reproduce insufflation, suction, and bowel wall motility. An ASGE technology assessment statement on simulators describes in detail the innovative technological developments in this field [102]. The images on the display can be derived from interactive video stored on laser disk, computer-generated images, or a combination of both.

At present, two commercially available simulators have reached the market. Other simulators described in the literature or in development, but not yet available publicly, will not be discussed in this chapter.

AccuTouch®   

The AccuTouch® Endoscopy Simulator (Immersion Medical) system simulator (Fig. 15) (http://www.immersion.com/products/medical/endoscopy.shtml) allows training in a number of procedures, including bronchoscopy, flexible sigmoidoscopy, and colonoscopy. It is possible to practice mucosal biopsy on this model. Direct performance feedback is provided by the simulator to the trainee. A number of preliminary validation studies are underway using this simulator, described in a recent editorial on simulators and presented in abstract form [112–115].

A small study by Tuggy compared five family practice residents who underwent simulator training to five others without training as they performed a real sigmoidoscopy on the same patients; after 6–10 h of simulator training, residents had improved insertion times and hand–eye skills measures (directional errors, percent of lumen visualized [114]. Gerson and Van Dam compared the performance on five actual sigmoidoscopies of seven internal medicine residents who received standard instruction on 10 patients to that of nine residents who received only unlimited training time on the AccuTouch® Simulator. Superior results were achieved by trainees in the standard teaching arm. However, the mean training time on the simulator was only 138 min, and it remains possible that either more extensive simulator work or a combination of simulator with real case experience will provide improvement over standard training methods [113].

If computer simulators are to have a role in credentialing in addition to training, they must be able to distinguish between a novice and an accomplished endoscopist. A recent study from the Mayo Clinic demonstrated that performance parameters on the simulator vary according to real colonoscopy experience. Internal medicine residents without prior experience, second-year GI fellows, and staff gastroenterologists each performed two simulated colonoscopies.

Significant differences were observed between residents and fellows and between faculty and fellows in terms of red out time, total procedure time, and insertion time, but not percent of mucosa seen, depth of insertion, or estimation of patient discomfort. While faculty rated the simulator endoscope handling and graphics as very realistic, the mean procedure time for faculty was 7 min, compared to an 18 min real colonoscopy average procedure time, suggesting that the degree of difficulty on the simulator is lower than it is on real procedures [115].

GI Mentor™   

The GI Mentor (Simbionix) (Figs 16–20) (http://www.simbionix.com/GI_Mentor.html) offers several diagnostic and therapeutic modules [108]. Upper and lower endoscopy, ERCP, and EUS are all performed on the same mannequin using a special endoscope for each procedure type. An accessory channel allows the endoscopist to perform a variety of therapeutic techniques including biopsy, polypectomy, sclerotherapy, and electrocoagulation to control active bleeding, ERCP cannulation, and sphincterotomy. This simulator also includes some manual dexterity training exercises, ideal for beginners to use to develop skills in controlling the endoscope dials and using torque. The logical descendant of the Lucero model and progressive training program, the simulator incorporates a series of cases of varying pathology and technical difficulty.

Specific training programs may be delineated by instructors. Trainees can get immediate feedback during and after completing each simulated procedure. In fact, the computer will even generate an expression of pain for over-insufflation or excessive looping of the instrument. Performance is recorded, including numbers and types of errors made. The instructor can review the progress of each trainee and the written procedure reports to determine whether abnormalities were correctly detected and identified; feedback messages may be sent back to the trainee.

The Simbionix computer simulator has been incorporated into a number of European endoscopy courses, most notably in Scandinavia [116–118]. Respondents to questionnaires have expressed great satisfaction with their limited experience on the Simbionix simulator. However, more objective validation data are currently being obtained.

In a small randomized study of 13 novices, in which half of the trainees received 2 h a day lower GI simulator training for 3 weeks, significant differences in the simulator group were observed using performance on the simulator to measure baseline and follow-up skills levels. Improvement was manifested by faster insertion times, time of unrestrained lumen view, and ability to pop a bubble with a needle accessory during a virtual dexterity test. The novice group also exhibited significantly less skill on baseline testing than a group of 11 expert endoscopists evaluated on the simulator performing the same simulated procedures [119]. Studies involving larger numbers of trainees and using performance on real colonoscopies as endpoints will be required to better assess the efficacy of this innovative technology.

In another trial, 20 beginners were randomized to 10 h on the upper GI Mentor model vs. no training and compared on their performance of their first 20 real EGDs; the simulator group had significantly fewer incomplete procedures, required less assistance, and exhibited better cognitive skills than the untrained group [120].

Current status of simulators   

It is impossible at this time to draw any conclusions from the differences in benefit observed between this trial and the flexible sigmoidoscopy validation study described above. Even if subsequent data demonstrate benefit from working on these two simulators, many additional questions will need to be addressed. Defining the optimal amount and frequency of simulator training prior to beginning real supervised cases will be essential. Will the impact on training justify the significant cost involved with providing access to computer endoscopy simulators? Another open question is whether simulator training will affect only the initial phase of endoscopy learning. Even if initial benefit is limited to the first 25 real procedures for upper or lower endoscopy, the impact of prior simulator experience on patient satisfaction during the initial stages of training will also need to be determined.

At present, computer simulators appear to have a lot to offer trainees in terms of showing diverse pathology and teaching beginners hand–eye coordination and scope handling. Unique aspects of this type of training are the simulation of contractions, feedback on comfort, opportunity for self-instruction without constant expert supervision, quantification of skills, and offsite skills assessment by instructors. Current available models appear less useful for more experienced endoscopists, although capabilities are expanding rapidly. At present the therapeutic modules for the Simbionix simulator are best suited for introductory orientation to polypectomy, hemostasis, and ERCP. An EUS module is the most recently released application.

Computer technology offers the potential to incorporate didactic lessons, specific questions for the endoscopist concerning accessory set-up and generator settings, and opportunities for self-assessment quizzes to complement the hands-on technical experience. The advent of forced feedback within the accessories will further enhance the realism of therapeutic endoscopy simulation.

Costs of simulators   

The major obstacle to expanded use of these simulators remains the cost and logistics of making them accessible to trainees. At costs ranging from $20 000 to $50 000, most individual departments cannot afford to purchase computer simulators. Potential solutions to the feasibility problem currently limiting dissemination of this technology include regional shared facilities, use of Web-based servers to transmit simulator modules to remote sites via broadband access, and the addition of simulator modules for other specialties so that hospitals could purchase one unit for use in various departments.

EUS models and simulators   

Previously, EUS training tools were limited to anatomical CD-ROM and live-animal course format. At present a variety of new hands-on training methods are under development. These range from simple static models to combined static–animal tissue models, to computer simulation. A rectangular gelatin-based container with embedded grapes can be used by beginners to approximate the technique of FNA. More realistic models incorporating ex vivo tissue, and artificial simulation of lymph nodes and flowing blood through the great vessels, are under development. EUS training, by virtue of the need to master the anatomical relationships in health and disease in many planes, is perhaps the area in which the virtual reality simulators will prove the most useful. Currently available using an echoendoscope in a mannequin, a lower-tech simulation using a PC with mouse for private training is under development, which might provide less expensive and more easily disseminated training opportunities [Yang Chen, personal communication].

Use of training resources: summary   

A possible algorithm for the incorporation of the various learning modalities available to complement supervised procedures in the process of endoscopy education is shown in Fig. 21.

 Endoscopy training 2010—a glimpse into the future   

• Computer simulators are used to test innate hand–eye coordination skills of fellowship applicants.
  
• A standard introduction to an endoscopy self-instruction didactic module must be completed over the Web by trainees before supervised instruction begins. Self-assessment video quizzes and ask-the-expert feedback are incorporated into this program.
  
• Training programs will offer static mannequins to allow novices to practice rudimentary controls on the endoscopes.
  
• Programs with sufficient resources will provide access to hospital-based virtual reality simulators which are designed to offer training in many GI and non-GI procedures. Hospitals will purchase these for training and credentialing of practitioners in many fields, and for the training of technical assistants for these procedures.
  
• Regional Web-based virtual reality servers will allow programs to subscribe and perform specific procedures on 'dummy' terminals at remote sites without purchasing the entire computer and software packages.
  
• Simulator work will allow manual dexterity training and supplement the Web-based tutorial as an introduction to pathology.
  
• Assessment capabilities on the simulators will indicate when trainees are ready to proceed to perform supervised real cases.
  
• All real cases will require computerized report documentation which will detail successful completion of intended endpoints. This will serve to facilitate calculation of success rates as objective measures of competency for fellows and practitioners alike.
  
• Central collection of endoscopy outcome data, such as currently performed using the CORI project, will become mandated by third-party payers and government agencies, of course keeping patient-identifying information confidential. This will have the required adoption of universal report generation using common terminology and compatible software.
  
• Advanced simulators will be mass-produced to allow for frequent regional workshops. Practicing endoscopists will be required to attend one such workshop every 5 years to maintain privileges for therapeutic endoscopy. Courses will include hands-on training, introduction to new techniques, and objective skills evaluation. GI fellows will be required to attend one such workshop at the start of the second year of fellowship, with the recommendation that they attend another during the third year. A cadre of endoscopy instructors, including individuals from every region of the country, will be trained on running these workshops. A full-time team hired by the ASGE with industry funding will be responsible for setting up and running the workshops at rotating hospitals throughout the country. Fellows will attend free of charge, and practicing endoscopists' fees will defray costs.
  
• Live-animal work will be reserved to research protocols designed to test the physiological and histopathological effects of new interventions.
  
• After much contention, GI, surgical, and other professional societies will agree on common outcome-based standards for credentialing in specific endoscopic procedures which will be similar regardless of the practitioners' specialty. Objective performance data available for each case done by a trainee will allow ready inquiry as to whether predetermined minimal acceptable levels of performance have been met.
  
• Non-medical personnel will be trained in similar fashion to perform various endoscopic procedures under the direction of on-site physician endoscopy specialists. For example the endoscopist would only actually perform the polypectomies or scope withdrawal from the cecum.
  
• Software embedded in the video endoscopic processor will be developed to recognize abnormal pathology and serve as a double check to alert technician and supervising endoscopist, to prevent missed lesion recognition.
  
• New techniques will be tested first on animal tissue models. Part of this testing would require analysis of the learning curve on the model. This information would serve as a guide for practitioners wanting to perform the new technique.
  
• Availability of ready hands-on training, even beyond the fellowship setting, and mandates for hands-on experience and supervised instruction for all new procedures from insurers and hospitals, will end the era of self-taught endoscopy.
  

 Credentialing and granting of privileges   

The process of privileging and credentialing endoscopists is designed to maximize patient safety and outcomes by insuring that individuals performing procedures are sufficiently competent to perform them at the time the privileges are being requested. Decisions by institutions about whom to privilege for which procedures and about whom to grant renewal of privileges depend on both the credentials of an applicant and documentation of current ability to perform the requested procedures at a predetermined acceptable level of success.

Credentialing   

Credentialing is the review of evidence that a prospective endoscopist has proper licensure, education, and adequate training to qualify them for privileges at an institution. Training directors are asked to attest to the competence level of trainees after completion of the training program. This certification refers to both technical and cognitive skill levels attained by the trainee. Trainees' procedure logs also provide documentation of the number and type of procedures performed.

Such information is used by institutions to help determine whether to grant privileges to individuals. Information regarding the number of procedures a fellow actually performed without assistance, the difficulty level of cases done, and the success rates achieved, is not generally available for either the training directors or for subsequent review by credentialing committees.

Other relevant data provided to institutional credentialing committees may include licensing information, specialty board certification, and documentation of attendance at postgraduate courses and other CME programs. Statements written by endoscopy teachers may also serve to support a trainee's application for privileges at a particular institution.

As far as the level of subspecialty board certification in gastroenterology, endoscopic interpretation of findings, indications, and management decisions are certainly covered on standardized tests. However, no practical demonstration of endoscopic skill is incorporated into credentialing examinations.

Privileging   

Similar to assessment of competency discussed above, review of credentials and determination of privileging requests should follow a few basic principles. These are specified in guidelines for credentialing and granting privileges published by the ASGE [10,69]. First, each procedure should be considered separately. Second, besides flexible sigmoidoscopy, individuals seeking privileges for diagnostic endoscopy should have the ability, and should seek the privileges, to perform those common therapeutic maneuvers which are typically encountered.

As discussed above, one should not grant colonoscopy privileges to someone who cannot perform—and provide the credentials to support that they can perform—polypectomy. Applicants must be able to document completion of appropriate forms of training and attainment of sufficient levels of both technical and cognitive competence. As with guidelines about competence, privileges guidelines stress the fact that minimum numbers of procedures performed does not ensure competency.

Proctoring   

Most information available to privileging boards is either indirect evidence of technical ability or written indication of cognitive skill. No separate standardized endoscopy certification examination or practical demonstration of skill on patients or simulators is presently practiced. The feasibility and acceptability of such certification, and the demand for it, are all open questions at this time. Direct demonstration of technical proficiency is possible at local institutions. It has been advocated that proctoring be conducted for all endoscopic procedures [121–124]. However, formal proctoring of individuals applying for privileges is not yet widely practiced.

ASGE guidelines   

The ASGE has published specific guidelines for hospitals intending to establish proctoring programs [125]. Proctoring protocols may be established for first-time applicants, for practicing gastroenterologists wishing to perform new techniques, for routine recredentialing purposes, and even for quality assurance purposes when potential problems are identified in a hospital unit. Proctors serve on behalf of the institution and must themselves be experienced endoscopists with expertise in the procedures they are asked to evaluate.

Ideally, the institution will invite outside experts to serve as proctors when no one already at the institution is qualified to do so for a particular procedure. While the guidelines state that proctors should be free of conflict of interest, this may be difficult to achieve in practice. Policies regarding proctoring and the implications of a failure to demonstrate acceptable levels of competency during proctored examinations need to be clearly delineated and written into hospital bylaws. For this process to work, the governing board in charge of privileging must determine an appropriate number of cases for each procedure type to be observed over a specified time period.

Standard evaluation forms for each major procedure should be submitted along with confidential written assessments to the governing body for decisions regarding privileging. Appeal mechanisms for applicants must be specified. The guidelines emphasize that proctors are not supposed to have any role in the actual procedures in terms of either patient contact or advice given to the applicant. Proctors should only intervene in the event of observed substandard care with potential immediate harm to patients. Complete documentation of proctoring activities is essential to minimize liability. Patients do need to be informed of the proctors' presence and role, and this also needs to be recorded [125,126].

 Renewal of privileges and privileging in new procedures   

The ASGE has published guidelines on the process of privilege renewal [127] (http://www.asge.org/gui/resources/manual/pc_maintaining.asp). Recredentialing of practitioners by hospitals is required at least every 2 years, as stipulated in the Joint Commission (JCAHO) Comprehensive Accreditation Manual for Hospitals [128]. Renewal of privileges depends on documentation of the following:

• Case volume sufficient to maintain skills.
  
• Objective data on procedure success rates and complications.
  
• Participation in relevant continuing education activities.
  

Data on the number of cases performed are perhaps the most easily obtained for the purpose of privilege renewal. As discussed above, there is some evidence that higher procedure volume leads to better patient outcomes and fewer complications. Individuals with very little experience in certain procedures may elect not to seek renewal for one or more technique. Just how many cases per year an endoscopist must perform to maintain acceptable quality of care is poorly characterized for most procedures [31].

In the absence of such data, those required to grant privilege renewals are left to make their own determinations as to whether documented procedure volume is adequate to maintain skills. If an endoscopist has not performed enough cases to maintain skills, a privileging board can recommend that the practitioner seek some additional training. The gold standard in the privilege renewal process really is procedure outcome data. This sort of objective information should become more readily available to privileging boards as they gain improved access to it under the auspices of quality assurance efforts. CME activity is expected of individuals applying for privilege renewal.

But how effective are refresher courses - or even intensive hands-on simulator tutorials - in skills maintenance? If these are effective tools in this regard, how often should an endoscopist participate in such activities? The answer would clearly vary with an individual's actual case volume and total prior experience. At present there are few objective data to address these important questions.

New procedures   

New procedures present an interesting problem for credentialing committees. For these techniques, requesting physicians are unlikely to have received much formal mentored training and little is known about the learning curve for the procedure [129]. There may be great variance in the overall endoscopic expertise and experience of individuals seeking privileges. For these reasons, some objective criteria of competency and some formal process of proctoring are particularly important in the process of granting privileges in new techniques [69,130].

An important distinction must be made between major new procedures, such as EUS, requiring formal preceptorship, and minor new techniques which might be adequately learned from didactic materials and attendance at short courses. An example of a minor new skill deemed to be an extension of normal practice would be the use of a new accessory.

 Privileging for non-gastroenterologists and non-physician providers   

Increasing demands for endoscopic procedures provide impetus for non-gastroenterologists and even non-physicians to perform certain endoscopic procedures. One advantage of formal, standardized institutional privileging processes which base decisions on objective documentation of competency and outcome is the common approach to prospective endoscopists regardless of specialty.

Recommendations for training vary widely among different specialties, as described above. Common privileging standards based on objective criteria would circumvent any disagreement by different specialties about minimum required training. For example, if a hospital credentialing and privileging board decides that applicants should have completed a minimum of 100 colonoscopies in training and be able to documentat least a 90% cecal intubation rate to merit privileges in colonoscopy, these would apply to gastroenterologists, surgeons, and family practitioners alike.

Such uniformity of standards would serve to minimize bias as to which providers are allowed to perform which procedures. Interestingly, in many instances in which paramedical personnel perform screening flexible sigmoidoscopy, the criteria used to privilege these individuals are actually more stringent than those applied to physicians [33].

One interesting wrinkle to the credentialing and privileging issue is the emerging practice of office-based endoscopy. With current reimbursement forces creating large sites-of-service differentials in provider fees, the prospect of increased performance of endoscopy in the unregulated office settings poses questions pertaining to training in addition to the obvious questions regarding patient safety implications. The ASGE guidelines for privileging and credentialing are intended to cover endoscopy performed by any provider regardless of specialty and regardless of setting. While ethical and medico-legal considerations should serve to prevent unqualified individuals from performing office-based endoscopic procedures, this remains a less regulated environment than the hospital.

 The future of credentialing and privileging   

While it is hard to predict how the process will change in the coming years, it is likely that standard guidelines will be more widely adopted. Local variation in practice may diminish with greater acceptance of these guidelines. Formal proctoring programs as described above may become more common. It is uncertain how government regulation and third-party payers will influence the process.

Two themes which appear in most published statements on competency and granting of privileges are an increasing reliance on objective measurement of competency to establish credentials and the importance of procedure outcome data. As such data become better collected and more readily available to hospital privileging boards, they will certainly have increasing importance in the privileging process and the renewal of privileges.

The use of new technology for credentialing   

The role of new technology for skills assessment has the potential to transform this process, but this remains problematic for a number of reasons. Simulators have the potential for current skills demonstration without patient risk and in a time-efficient manner [131]. However, access to simulators with proctors to evaluate skills on them is quite limited at present. As described above, preliminary data are currently being collected to validate the utility of various simulators as effective learning tools.

Another key question in the simulator field is the applicability of skills demonstration on simulators to performance on real endoscopy. If good correlation of performance evaluations on computer simulators and animal-tissue models with proctored evaluations on real examinations can be demonstrated, there will be a great impetus for wider dissemination of simulators.

Hospitals may also feel impelled to financially support the purchase of simulators to facilitate the credentialing and privileging processes. Assuming that simulators can now or one day provide an accurate means of testing endoscopic skill, the same cost and logistical concerns described above pertaining to their use in training will need to be addressed before they are adopted for general use in credentialing.

 The role of endoscopic societies in training and credentialing   

Professional organizations and societies have a great impact in shaping the content and conduct of endoscopy training and credentialing.

Guidelines   

First of all, guidelines set forth at a national level influence the community standard of care. Individual training programs look to such recommendations when reviewing themselves. Standards of practice in training are also used by accreditation bodies which conduct regular reviews of each US residency and fellowship program. Beyond this, efforts to create a more standard endoscopy training curriculum may offer more specific guidance to any future program restructuring.

For example, the ASGE Training Committee has expressed interest in establishing a common introduction to the endoscopy curriculum for incoming fellows. A program along these lines has been pioneered by the New York Society for Gastrointestinal Endoscopy (NYSGE) (http://www.nysge.org). Originally an informal series of after-hours sessions in endoscopy units around the city, the NYSGE summer course now consists of one full weekend of didactic lectures given by regional experts for over 40 new area GI fellows and a similar weekend course for returning fellows on therapeutic procedures. The NYSGE makes the syllabus available to fellows and society members on CD-ROM.

Society courses   

Besides setting standards to guide local programs, the national organizations sponsor a number of projects which provide didactic and hands-on learning opportunities for trainees and practitioners. These range from the large multitopic postgraduate courses at national meetings to regional interim courses on a particular topic. The formats at these events vary, and may include combinations of annotated video presentations, live demonstrations, standard lectures, and, more recently, hands-on models.

Materials   

The ASGE has sponsored the creation of a wide array of instructional procedure videotapes, which can be purchased by individuals and by fellowship programs. More recently, this video material has been downloaded to the ASGE website; broadband streaming of annotated video or live Web-casts will likely greatly increase the reach of such efforts to propagate expert endoscopy instruction to supplement supervised endoscopic training at local institutions.

Hands-on courses   

Endoscopy societies also play an important role in making available hands-on training opportunities to trainees and members. In the US, the best examples have been the live-animal courses sponsored at the Ethicon Institute on such areas as ERCP and EUS. In Europe, the Scandinavian Association of Digestive Endoscopy (SADE) has made great strides to institute an integrated series of regional courses which use didactic presentations as well as simulators [116].

The German Endoscopy societies are preparing to devise a standard training curriculum which provides all trainees with the opportunity to get hands-on expert instruction on the porcine stomach, and artificial tissue simulators designed for this purpose. The NYSGE has incorporated such hands-on intensive workshops for registrants into its annual course; it has already used the EASIE model in the above-described study to train area fellows from multiple training programs in hemostasis techniques.

In an important development, the ASGE has established a permanent training facility at its headquarters in Chicago, with plans to conduct frequent training courses of fellows using all available simulators and hands-on models. In these ways, endoscopy societies may provide the organization, personnel, funding, and motivation to overcome the logistical obstacles which prevent more widespread access to simulator training opportunities.

Research in training   

Research in the area of training in endoscopy has produced much of the objective data and many of the innovative training methods presented above. Much work remains to be done both to validate new training modalities and to assess how patient outcomes are influenced by training. Whether by emphasizing the importance of work in this area, by directly funding training-related research projects, or by directly organizing collaborative projects in this area using their memberships, regional and national societies may impact substantially on future directions in this field.

Influencing credentialing   

Finally, the work of societies to establish guidelines for training and to publicize this information in journals and on the Web has obvious influence on local credentialing boards and hospital privileging committees. Apart from providing guidelines on training which local boards may use to assess the training experience of a particular applicant, the ASGE has separate policy statements available on credentialing and privileges (http://www.asge.org/gui/resources/manual/pc_index.asp).

However, many local institutions have not implemented formal protocols for assessing competency in specific procedures requested as privileges. With increasing scrutiny of hospital privileging practices by regulatory agencies, training directors and individuals responsible for credentialing may well look to the guidelines provided by endoscopy societies in updating their practice and policies.

 Outstanding issues and future trends   

The technology of endoscopy has been progressing rapidly, dramatically expanding the diagnostic and therapeutic potential of the procedures performed. Slowly, the training process is evolving, as it must, into a more controlled process which incorporates data concerning acquisition of competency.

Standard supervised practice on real patients will be supplemented by alternative training modalities at appropriate stages in the training process. This will begin with more didactic and interactive media presentations of technique and pathology recognition for novices along with simulators to teach hand–eye coordination and control of the instruments. Trainees will thereby begin real endoscopy with a running start, and free patients from the delays and potential discomfort of being subject to true beginners.

Hands-on training workshops for therapeutic techniques will enhance basic skills and teach specific procedures in order to provide opportunities for independent performance of techniques, repetitive practice of skills, and relaxed feedback not reliably available during the usual course of training.

Live courses and annotated video presentations will complement, but by no means substitute for, dedicated training in advanced procedures and skills. Short courses involving hands-on techniques may serve multiple purposes: as starting-off points for more concentrated supervised training, as intensive comprehensive training in certain specific new techniques such as the use of a new accessory, and as retraining opportunities for endoscopists. Increasingly, when fellows complete their endoscopy training, and when practitioners learn a new technique, they will be required to demonstrate evidence of satisfactory procedure success rates before getting privileged to practice these techniques without supervision.

Besides transforming the credentialing and privileging process, outcome data on procedure success, complications, and patient satisfaction will be the best validation for future efforts to improve the effectiveness and efficiency of endoscopy training.

 References  

  1 Sivak, MV. The art of endoscopic instruction. Gastrointest Endosc Clin N Am 1995; 5: 299–310. PubMed

  2 Church, JN. Learning colonoscopy: the need for patience (patients). Am J Gastroenterol 1993; 88: 1569.

  3 Katz, PO. Providing feedback. Gastrointest Endosc Clin N Am 1995; 5: 347–55. PubMed

  4 Borland, JL. Retraining in endoscopy. Gastrointest Endosc Clin N Am 1995; 5: 363–72. PubMed

  5 American Society for Gastrointestinal Endoscopy. Guidelines for clinical application: statement on role of short courses in endoscopic training. Gastrointest Endosc 1999; 50: 913–14. PubMed

  6 American Society for Gastrointestinal Endoscopy. Principles of training in gastrointestinal endoscopy. Gastrointest Endosc 1999; 49: 845–50. PubMed

  7 Cass, OW, Freeman, ML & Cohen, J et al. Acquisition of competency in endoscopic skills (ACES) during training: a multicenter study [Abstract]. Gastrointest Endosc 1996; 43: 308.

  8 Cass, OW, Freeman, ML & Peine, CJ et al. Objective evaluation of endoscopy skills during training. Ann Intern Med 1993; 118: 40–4. PubMed

 9 Jowell, PS, Baillie, J & Branch, S et al. Quantitative assessment of procedural competence: a prospective study of training in endoscopic retrograde cholangiopancreatography. Ann Intern Med 1996; 125: 983–9. PubMed

10 American Society for Gastrointestinal Endoscopy. Guidelines for credentialing and granting privileges for gastrointestinal endoscopy. Gastrointest Endosc 2002; 55: 780–3. PubMed

11 Coppola, AG, Raufman, JP & Sharma, VK. Unregulated gastrointestinal endoscopy credentialing practices: a national survey [Abstract]. Gastrointest Endosc 2001, 53: AB55.

12 Freeman, ML. Training and competence in gastrointestinal endoscopy. Rev Gastroenterol Disord 2001; 1: 73–86. PubMed

13 Wigton, RS, Nicolas, JOA & Blank, LL. Procedural skills of the general internists: a survey of 2500 physicians. Ann Intern Med 1989; 111: 1023–34. PubMed

14 Wigton, RS, Blank, LL & Nicolas, JOA. Procedural skills training in internal medicine residencies. Ann Intern Med 1989; 111: 932–8. PubMed

15 Wigton, RS, Blank, LL, Monsour, H & Nicolas, JOA. Procedural skills of practicing gastroenterologists: a national survey of 700 members of the American College of Physicians. Ann Intern Med 1990; 113: 540–6. PubMed

16 American Board of Internal Medicine. Results of procedure survey of gastroenterology program directors. Am Board Intern Medical Newsletter1990; Spring/Summer: 4–5.

17 Federation of Digestive Disease Societies. (1981) Guidelines for training in endoscopy. Manchester, MA. Federation of Digestive Disease Societies.

18 Society for American Gastrointestinal Surgeons. Granting privileges for gastrointestinal endoscopy by surgeons. SAGES publication no., 11. Los Angeles: Society for American Gastrointestinal Surgeons; 1991.

19 Rodney, WM, Hocutt, JE & Coleman, WH et al. Esophagogastroscopy by family physicians: a national multisite study of 717 procedures. J Am Board Fam Pract 1990; 3: 73–9. PubMed

20 European Union of Medical Specialists European Board of Gastroenterology 1995 Requirements for the specialty gastroenterology. In: Charter on training of medical specialists in the EU. Brussels: European Union of Medical Specialists, European Board of Gastroenterology; : Available from: URL http://www.uems.be/gastrointestinal-e.htm.

21 Farthing, MJG, Walt, RP & Allan, RN et al. A national training programme for gastroenterology and hepatology. Gut 1996; 38: 459–70. PubMed

22 Conjoint Committee for Recognition of Training in Gastrointestinal Endoscopy. (1997) Information for supervisors: changes to endoscopic training. Sydney: The Conjoint Committee for Recognition of Training in Gastrointestinal Endoscopy;

23 German Society of Surgery. Standards and guidelines of flexible surgical endoscopy: recommendations of the Surgical Study Group of Endoscopy and Sonography of the German Society of Surgery and the Working Group of German Surgeons on learning and quality assurance in intraluminal endoscopy in surgery. Chirurg 1999; 70: 245–6. PubMed

24 Rodney, WM, Weber, JR & Swedberg, JA et al. Esophagogastroduodenoscopy by family physicians Phase II: a national multisite study of 2,500 procedures. Fam Pract Res J 1993; 13: 121–31. PubMed

25 Schauer, PR, Schwesinger, WH & Page, CP et al. Complications of surgical endoscopy: a decade of experience from a surgical residency training program. Surg Endosc 1997; 11: 8–11. PubMed CrossRef

26 Sieg, A, Hachmoeller-Eisenbach, U & Eisenbach, T. Prospective evaluation of complications in outpatient GI endoscopy: a survey among German gastroenterologists. Gastrointest Endosc 2001; 53: 620–7. PubMed CrossRef

27 The Gastroenterology Leadership Council. Training the gastroenterologist of the future: the gastroenterology core curriculum. Gastroenterology 1996; 110: 1266–300. PubMed

28 American Society for Gastrointestinal Endoscopy. Principles of privileging and credentialing for endoscopy and colonoscopy: granting privileges for GI endoscopy. Gastrointest Endosc 2002; 55: 145–8. PubMed

29 Hochberger, J, Cohen, J, Matthes, K, Maiss, J, Koebnick, C & Hahn, E, New York Society for Gastrointestinal Endoscopy (NYSGE) Study Group. Randomized, prospective comparison of compact EASIE simulator hemostasis training with standard endoscopy education [Abstract]. Gastrointest Endosc 2002; 55: AB78.

30 Stevens, PD, Finegold, J & Garcia-Carrasquillo, RJ et al. Effectiveness of a protocol-based team approach to gastrointestinal hemorrhage [Abstract]. Gastrointest Endosc 1997, 45: AB102.

31 American Society for Gastrointestinal Endoscopy. Maintaining competency in endoscopic skills. Gastrointest Endosc 1995; 42: 620–1. PubMed

32 Hawes, R, Lehman, GA & Hast, J et al. Training resident physicians in fiberoptic sigmoidoscopy: how many supervised examinations are required to achieve competence? Am J Med 1986; 80: 465–70. PubMed CrossRef

33 Cash, BD, Schoenfeld, PS & Ransohoff, DF. Licensure, use, and training of paramedical personnel to perform screening flexible sigmoidoscopy. Gastrointest Endosc 1999; 49: 163–9. PubMed

34 Lieberman, DA, Weiss, DG & Bond, JH et al. Use of colonoscopy to screen asymptomatic adults for colorectal cancer. N Engl J Med 2000; 343: 162–8. PubMed CrossRef

35 Rex, DK, Rahmani, EY & Haseman, JH et al. Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology 1997; 112: 17–23. PubMed

36 Parry, BA & Williams, SM. Competency and the colonoscopists: a learning curve. Aust N Z J Surg 1991; 61: 419–22. PubMed

37 Marshall, JB. Technical proficiency of trainees performing colonoscopy: a learning curve. Gastrointest Endosc 1995; 42: 287–91. PubMed

38 Chak, A, Cooper, GS & Blades, EW et al. Prospective assessment of colonoscopic intubation skills in trainees. Gastrointest Endosc 1996; 44: 54–7. PubMed

39 Church, JN. 1995 Training. In: Endoscopy of the Colon, Rectum and Anus (ed Church JN). pp 214–25 Igaku Shoin, New York.

40 Tassios, PS, Ladus, SD & Grammenos, I et al. Acquisition of competence in colonoscopy: the learning curve of trainees. Endoscopy 1999; 31: 702–6. PubMed CrossRef

41 Cass, OW. Training to competence in gastrointestinal endoscopy: a plea for continuous measuring of objective end points. Endoscopy 1999; 31: 751–4. PubMed

42 Rodney, WM, Dabov, G & Cronin, C. Evolving colonoscopy skills in a rural family practice: the first 293 cases. Fam Pract Res J 1993, 1343–52.

43 Harper, MB, Pope, JB & Mayeaux, EJ et al. Colonoscopy experience at a family practice residency: a comparison to gastroenterology and general surgery services. Fam Med 1997; 29: 575–9. PubMed

44 Hopper, W, Kyker, KA & Rodney, WM. Colonoscopy by a family physician: a 9-year experience of 1048 procedures. J Fam Pract 1996; 43: 561–6. PubMed

45 Godreau, CJ. Office-based colonoscopy in a family practice. Fam Pract Res J 1992; 12: 313–20. PubMed

46 Haseman, JH, Lemmel, GT, Rahmani, EY & Rex, DK. Failure of colonoscopy to detect colorectal cancer. Gastrointest Endosc 1997; 45: 451–5. PubMed

47 Wexner, SD, Forde, KA & Sellers, G et al. How well can surgeons perform colonoscopy? Surg Endosc 1998; 12: 1410–14. PubMed CrossRef

48 Bond, JH & Frakes, JT. Who should perform colonoscopy? How much training is needed? Gastrointest Endosc 1999; 49: 657–9. PubMed

49 American Society for Gastrointestinal Endoscopy. Guidelines for advanced endoscopic training. Gastrointest Endosc 2001; 53: 846–8. PubMed

50 Freeman, ML, Nelson, DB & DiSario, JA et al. Risk factors for post ERCP pancreatitis: a prospective, multicenter study [Abstract]. Gastrointest Endosc 1999, 49: AB72.

51 Choudari, CP, Sherman, S & Fogel, EL et al. Success of ERCP at a referral center after a previously unsuccessful attempt. Gastrointest Endosc 2000; 52: 478–83. PubMed

52 Rollhauser, C, Benjamin, SB & Al-Kawas, FH. Success of ERCP at an academic center after referral for a failed cannulation or failure to complete therapeutic goal [Abstract]. Gastrointest Endosc 1997, 45: AB146.

53 Cunningham, JT, Tarnasky, PR, Hawes, RH & Cotton, PB. Repeat ERCP after prior failure can be safely performed with a high degree of diagnostic and therapeutic success [Abstract]. Am J Gastroenterol 1997, 92.

54 Tanner, AR. ERCP: present practice in a single region. Eur J Gastroenterol Hepatol 1996; 8: 145. PubMed

55 Schutz, SM & Abbott, RM. Grading ERCPs by degree of difficulty: a new concept to produce more meaningful outcome data. Gastrointest Endosc 2000; 51: 535–9. PubMed

56 Vennes, JA, Ament, M, Boyce, HW Jr, Cotton, PB, Jensen, DM & Ravich, WJ et al. Principles of training in gastrointestinal endoscopy. American Society for Gastrointestinal Endoscopy, Standard of Training Committees, 1989–90. Gastrointest Endosc 1992; 38: 743–7. PubMed

57 Watkins, JL, Etzkorn, KP, Wiley, TE, DeGuzman, L & Harig, JM. Assessment of technical competence during ERCP training. Gastrointest Endosc 1996; 44: 411–15. PubMed

58 Schlup, MMT, Williams, SM & Barbezat, GO. ERCP: a review of technical competency and workload in a small unit. Gastrointest Endosc 1997; 46: 48–52. PubMed

59 Davis, WZ, Cotton, PB & Arias, R et al. ERCP and sphincterotomy in the context of laparoscopic cholecystectomy: academic and community practice patterns and results. Am J Gastroenterol 1997; 92: 597–601. PubMed

60 Freeman, ML, Nelson, DB, Sherman, S, Haber, GB, Herman, ME & Dorsher, PJ et al. Complications of endoscopic biliary sphincterotomy. N Engl J Med 1996; 335: 909–18. PubMed CrossRef

61 Loperfido, S, Angelini, G & Benedetti, G et al. Major early complications from diagnostic and therapeutic ERCP: a prospective multicenter study. Gastrointest Endosc 1998; 48: 1–10. PubMed

62 Freeman, ML, Nelson, DB & Eisen, GM et al. Failures and complications of attempted therapeutic ERCP: impact on outcomes and costs [Abstract]. Gastrointest Endosc 1998, 47: AB114.

63 Freeman, ML. Procedure-specific outcomes assessment for endoscopic retrograde cholangiopancreatography. Gastrointest Endosc Clin N Am 1999; 9: 639–47. PubMed

64 Escourrou, J, Delvaux, M & Buscail, L et al. Clinical results of endoscopic sphincterotomy: comparison of 2 activity periods in the same endoscopy unit. Gastrointest Endosc 1990; 36: 205–6.

65 Rabenstein, T, Schneider, HT, Nicklas, M, Ruppert, T, Katalinic, A, Hahn, EG & Ell, C. Impact of skill and experience of the endoscopist on the outcome of endoscopic sphincterotomy techniques. Gastrointest Endosc 1999; 50: 628–36. PubMed

66 Bilbao, MK, Dotter, CT & Lee, TG et al. Complications of endoscopic retrograde cholangiopancreatography (ERCP). A study of 10,000 cases. Gastroenterology 1976; 70: 314–20. PubMed

67 Harewood, GC, Wright, CA & Baron, TH. Impact on patient outcomes of experience in the performance of endoscoit pancreatic fluid collection drainage. Gastrointest Endosc 2003; 58: 230–5. PubMed

68 American Society for Gastrointestinal Endoscopy. Guidelines for training in endoscopic ultrasound: guidelines for clinical application. Gastrointest Endosc 1999; 49: 829–33. PubMed

69 American Society for Gastrointestinal Endoscopy. Guidelines for credentialing and granting privileges for endoscopic ultrasound. Gastrointest Endosc 2001; 54: 811–14[http://www.asge.org/gui/resources/manual/pc_creditial_ultrasound.asp]. PubMed

70 Catalano, MF, Sivak, MV & Bedford, RA et al. Observer variation and reproducibility of endoscopic ultrasonography. Gastrointest Endosc 1995; 41: 115–20. PubMed

71 Hoffman, BJ, Wallace, MB & Bloubeidi, MA et al. How many supervised procedures does it take to become competent in EUS? Results of a multi-center three year study [Abstract]. Gastrointest Endosc 2000, 51 AB139.

72 Hoffman, BJ & Hawes, RH. Endoscopic ultrasound and clinical competence. Gastrointest Endosc Clin N Am 1995; 5: 879–84. PubMed

73 Fockens, P, Vanden Brande, JH & Van Dullemen, HA et al. Endosonographic T-staging of esophageal carcinoma: a learning curve. Gastrointest Endosc 1996; 44: 58–62. PubMed

74 Schlick, T, Heintz, A & Junginger, T. The examiner's learning effect and its influence on the quality of endoscopic ultrasonography in carcinoma of the esophagus and gastric cardia. Surg Endosc 1999; 13: 894–8. PubMed CrossRef

75 Boyce, HW. Training in endoscopic ultrasonography. Gastrointest Endosc 1996; 43: S12–5. PubMed

76 Harewood, GC, Wiersema, LM, Halling, AC, Keeney, GL, Salamao, DR & Wiersema, MJ. Influence of EUS training and pathology interpretation on accuracy of EUS-guided fine needle aspiration of pancreatic masses. Gastrointest Endosc 2002; 55: 669–73. PubMed CrossRef

77 Mertz, H & Personal Communication., GIE. In press.

78 Laine, L, Freeman, M & Cohen, H. Lack of uniformity in evaluation of endoscopic prognostic features of bleeding ulcers. Gastrointest Endosc 1994; 40: 411–17. PubMed

79 Cotton, PB. Live endoscopy demonstrations are great, but ... . Gastrointest Endosc 2000; 51: 627–9. PubMed CrossRef

80 Waye, JD, Axon, A, Riemann, JF & Chung, S. Continuing education in endoscopy: live courses or video format? Gastrointest Endosc 2000; 52: 447–51. CrossRef

81 Carr-Locke, DL, Gostout, CJ & Van Dam, J. A guideline for live endoscopy courses: an ASGE white paper. Gastrointest Endosc 2001; 53: 685–8. PubMed

82 Schindler, R. 1937Gastroscopy: the endoscopic study of gastric pathology. Chicago: University of Chicago Press, 74–5.

83 Heinkel, VK & Kimmig, JM. Megenphantome zur ausbiling in der gastrokamera-magenuntersuchung. Z Gastroenterol 1971; 9: 331. PubMed

84 Classen, M & Ruppin, H. Practical training using a new gastrointestinal phantom. Endoscopy 1974; 6: 127–31.

85 Williams, CB, Saunders, BP & Bladen, JS. Development of colonoscopy teaching simulation. Endoscopy 2000; 32: 901–5. PubMed CrossRef

86 Leung, JW & Chung, RS. Training in ERCP [Letter]. Gastrointest Endosc 1992; 38: 517. PubMed

87 Lucero, RS, Zarate, JO, Espinella, F, Davolos, J, Apud, A, Gonzalez, B & Grosso, M. Introducing digestive endoscopy with the 'SimPrac-EDF y VEE' simulator, other organ models and mannequins: teaching experience in 21 courses attended by 422 physicians. Endoscopy 1995; 27: 93–100. PubMed

88 Grund, KE, Bräutigam, D & Zindel, C et al. Interventionsfähiges Tübinger Simulationsmodell INTERPHANT für die flexible Endoskopie. Endoskopie Heute 1998; 11: 134.

89 Hochberger, J, Neumann, M, Hohenberger, W & Hahn, EG. Neues Bio-Trainingsmodell für die operative flexible Endoskopie. Endoskopie Heute 1997; 1: 117–18.

90 Hochberger, J, Neumann, M, Hohenberger, W & Hahn, EG. Neuer Endoskopie-Trainer für die therapeutische flexible Endoskopie. Z Gastroenterol 1997; 35: 722–3.

91 Hochberger, J, Maiss, J, Hahn, EG., Bhutani, MS & Tandon, RK, eds (2001). Advances in Gastrointestinal Endoscopy. New Delhi: Jaypee Brothers. Medical Publishers P. Ltd.

92 Hochberger, J, Maiss, J, Magdeburg, B, Cohen, J & Hahn, EG. Training simulators and education in gastrointestinal endoscopy: current status and perspectives in 2001. Endoscopy 2001; 33: 541–9. PubMed CrossRef

93 Neumann, M, Mayer, G, Ell, C, Felzmann, T, Reingruber, B & Horbach, T et al. The Erlangen Endo-Trainer: life-like simulation for diagnostic and interventional endoscopic retrograde cholangiography. Endoscopy 2000; 32 (11): 906–10. PubMed CrossRef

94 Hochberger, J, Maiss, J, Nägel, A, Tex, S & Hahn, EG. Polypectomy / vital endoscopic staining / mucosectomy: a new structured team–training in a close to reality endoscopy simulator (EASIE). Endoscopy 2000; 32 (Suppl. 1): E23.

95 Hochberger, J, Maiss, J, Neumann, M, Hildebrand, V, Bayer, J & Hahn, EG. EASIE-team-training in endoscopic hemostasis: acceptance of a systematic training in interventional endoscopy by 134 trainees [Abstract]. Gastrointest Endosc 1999, 49: AB143.

96 Hochberger, J, Neumann, M, Maiss, J, Hohenberger, W & Hahn, EG. EASIE—Erlangen Active Simulator for Interventional Endoscopy—a new bio-simulation-model: first experiences gained in training workshops [Abstract]. Gastrointest Endosc 1998, 47: AB116.

97 Gholson, CF, Provenza, JM, Silver, RC & Bacon, BR. Endoscopic retrograde cholangiography in the swine: a new model for endoscopic training and hepatobiliary research. Gastrointest Endosc 1990; 36: 600–3. PubMed

98 Falkenstein, DB, Abrams, RM, Kessler, RE, Jones, B, Johnson, G & Zimmon, DS. Endoscopic retrograde cholangiopancreatography in the dog: a model for training and research. Gastrointest Endosc 1974; 21: 25–6. PubMed

99 Noar, MD. An established porcine model for animate training in diagnostic and therapeutic ERCP. Endoscopy 1995; 27: 77–80. PubMed

100 Bhutani, MS, Hoffman, BJ & Hawes, RH. A swine model for teaching endoscopic ultrasound (EUS) imaging and intervention under EUS guidance. Endoscopy 1998; 30: 605–9. PubMed

101 Pasricha, PJ, Tietjen, TG & Kalloo, AN. Biliary manometry in swine: a unique endoscopic model. Endoscopy 1995; 27: 70–2. PubMed

102 American Society for Gastrointestinal Endoscopy. Technology status evaluation report: endoscopy simulators. Gastrointest Endosc 1999; 50: 935–7. PubMed

103 Williams, CB, Baillie, J, Gillies, DF, Borislow, D & Cotton, PB. Teaching gastrointestinal endoscopy by computer simulation. a prototype for colonoscopy and ERCP. Gastrointest Endosc 1990; 36: 49–54. PubMed

104 Noar, MD. Robotics interactive endoscopy simulation of ERCP/sphincterotomy and EGD. Endoscopy Supplement 1992; 2: 539–41.

105 Noar, MD. The next generation of endoscopy simulation: minimally invasive surgical skills simulation. Endoscopy 1995; 27: 81–5. PubMed

106 Noar, MD & Soehendra, N. Endoscopy simulation training devices. Endoscopy 1992; 24: 159–66. PubMed

107 Beer-Gabel, M, Delmontte, S & Muntlak, L. Computer assisted training in endoscopy (C.A.T.E): from a simulator to a learning station. Endoscopy 1992; 24 (Suppl. 2): 534–8. PubMed

108 Bar-Meir, S. A new endoscopic simulator. Endoscopy 2000; 32 (11): 898–900. PubMed

109 Baillie, J & Jowell, P. ERCP training in the 1990s: time for new ideas. Gastrointest Endosc Clin N Am 1994; 4 (2): 409–21. PubMed

110 Gessner, CE, Jowell, PS & Baillie, J. Novel methods for endoscopic training. Gastrointest Endosc Clin N Am 1995; 5: 323–36. PubMed

111 Bar-Meir, S. Endoscopy simulators: the state of the art, 2000. Gastrointest Endosc 2000; 52: 201–3.

112 Gerson, LB & Van Dam, J. The future of simulators in GI endoscopy: an unlikely possibility or a virtual reality? [Editorial.] Gastrointest Endosc 2002; 55: 608–11. PubMed CrossRef

113 Gerson, LB & Van Dam, J. A randomized controlled trial comparing an endoscopic simulator to traditional bedside teaching for training in flexible sigmoidoscopy [Abstract]. Gastrointest Endosc 2002, 55: AB78.

114 Tuggy, ML. Virtual reality flexible sigmoidoscopy simulator training: impact on resident performance. J Am Board Fam Pract 1998; 11: 426–33. PubMed

115 Sedlack, RE & Kolars, JC. Validation of computer-based endoscopy simulators in training [Abstract]. Gastrointest Endosc 2002, 55: ABAB77.

116 Aabakken, L, Osnes, M, Rosseland, AR, Fork, FT, Liedberg, G, Matzen, P & Kruse, A. Hands-on endoscopy training: an evaluation of the SADE endoscopy course. Endoscopy 1995; 27: 66–9. PubMed

117 Aabakken, L, Adamsen, S & Kruse, A. Performance of a colonoscopy simulator: experience from a hands-on endoscopy course. Endoscopy 2000; 32 (11): 911–13. PubMed CrossRef

118 Adamsen, S. Simulators and gastrointestinal endoscopy training. Endoscopy 2000; 32: 895–7. PubMed CrossRef

119 Ferlitsch, A, Glauninger, P, Gupper, A, Schillinger, M, Heafner, AG, Muellner, M & Schoefl, R. Virtual endoscopy simulation for training of gastrointestinal endoscopy [Abstract]. Gastrointest Endosc 2001, 53 AB78.

120 Fregonese, D, Cassetti, T, Cestari, R, Chilovi, F, D'Ambra, G & Delle Fave, G et al. Basic endoscopy training: usefulness of a computer-based simulator. Cooperative group for training in endoscopy [Abstract]. Gastrointest Endosc 2001, 53: AB81.

121 Health and Public Policy Committee, American College of Physicians. Clinical competence in the use of flexible sigmoidoscopy for screening purposes. Ann Intern Med 1987; 107: 589–91. PubMed

122 Health and Public Policy Committee, American College of Physicians. Clinical competence in diagnostic esophagogastroduodenoscopy. Ann Intern Med 1987; 107: 937–9. PubMed

123 Health and Public Policy Committee, American College of Physicians. Clinical competence in colonoscopy. Ann Intern Med 1987; 107: 772–4. PubMed

124 Health and Public Policy Committee, American College of Physicians. Clinical competence in diagnostic ERCP. Ann Intern Med 1988; 108: 142–4. PubMed

125 American Society for Gastrointestinal Endoscopy. Proctoring and hospital endoscopy privileges guidelines for clinical application. Gastrointest Endosc 1999; 50: 901–5. PubMed

126 Sundermeyer, M. Prevention of hospital liability for granting privileges to unqualified physicians. Gastrointest Endosc Clin N Am 1995; 5: 433–5. PubMed

127 American Society for Gastrointestinal Endoscopy. Renewal of endoscopic privileges: guidelines for clinical application. Gastrointest Endosc 1999; 49: 823–5. PubMed

128 Joint Commission on Accreditation of Healthcare Organizations. Comprehensive Accreditation Manual for Hospitals. Jt Comm Perspect 1997; 17(1): EC7–9. PubMed

129 Fleischer, DE. Advanced training in endoscopy. Gastrointest Endosc Clin N Am 1995; 5: 311–22. PubMed

130 American Society for Gastrointestinal Endoscopy. Guidelines for clinical application: methods of privileging for new technology in gastrointestinal endoscopy. Gastrointest Endosc 1999; 50: 899–900. PubMed

131 Rodney, WM. Will virtual reality simulators end the credentialing arms race in gastrointestinal endoscopy or the need for family physician faculty with endoscopic skills? J Am Board Fam Pract 1998; 11: 492–6. PubMed

Copyright © Blackwell Publishing, 2004