1.2.1 Clinical Training to Competency in Esophagogastroduodenoscopy and Colonoscopy: Studies, Guidelines, and Assessment

Since the early 1980s, trainees have been required to keep a record of all procedures performed, ¹ in particular for colonoscopy. The ability to reach the cecum is the most common criterion by which colonoscopies have been judged. ⁹ Data from early studies showed variable learning curves and led to the concept of minimal numbers of procedures required. ¹ Sedlack et al presented in 2011 a new assessment tool, the so-called Mayo Colonoscopy Skills Assessment Tool (MCSAT), to describe learning curves for colonoscopy. ¹⁰ They evaluated forty-one GI fellows who performed 6,635 colonoscopies. Independent cecal intubation rates of 85% and cecal intubation times of 16 minutes or less were achieved at 275 procedures on average, which is more than previous gastroenterology training recommendations required.

In 2014, the Training Committee of the American Society for Gastrointestinal Endoscopy (ASGE) presented the “Assessment of Competency in Endoscopy” (ACE) tool as a refinement of the MCSAT. ⁹ The ACE tool added important quality parameters such as a metric assessment of fine-tip control and polyp detection rates. In 2016, a prospective, multicenter trial was published evaluating the ACE tool at 10 institutions across the United States including 93 gastrointestinal (GI) fellows. ¹¹ A total of 184 senior endoscopists assessed 1,061 colonoscopies, which included 6 motor and 6 cognitive skills on a 4-point scale. The average fellow reached required cognitive and motor skills endpoints by 250 procedures, with over 90% of fellows surpassing these thresholds by 300 procedures. ¹¹ Procedure times, polyp detection rates, and polyp miss rates with increasing experience are shown in ▶Fig. 1.1 and ▶Fig. 1.2.

Barton et al ¹² described in 2012 the value of the Direct Observation of Procedural Skills (DOPS) method developed by an expert group of colonoscopists and clinical educators in the United Kingdom. Colonoscopists wishing to participate in the British National Health Service National Bowel Cancer Screening Programme (BCSP) were assessed. Assessments from 147 candidates and 28 assessors were analyzed. Candidates had to prove experience in a minimum of 500 colonoscopies with a self-reported cecal intubation rate of ≥ 90% and a polyp detection rate of ≥ 20%. The assessment had high reliability using generalizability theory (G) with G = 0.81 and correlated highly with a global expert assessment. Both, candidates and assessors, believed that the DOPS was a valid assessment of competence.

Anderson ¹³ recently described how DOPS evaluation has been successfully integrated for trainees as well as for independent endoscopists into the “UK National Bowel Cancer Screening Programme.” The Joint Advisory Group (JAG) sets the standards for endoscopy training and the accreditation of endoscopy units as base training units. ¹⁴ , ¹⁵ A Global Rating Scale web-based system is used for continuous assessment of performance and DOPS is regularly applied in order to monitor continuously individual performances. An individual web-based logbook and e-portfolio of each endoscopist is created via a national database system that is the base for credentialing and certification. Feedback of data to individuals helps in benchmarking and identification of those with suboptimal performance and a need for extra training and close audits. The system has recently been extended to upper GI endoscopy and other techniques. ¹⁶

1.2.2 Training in Endoscopic Retrograde Cholangiopancreatography

Proficiency in all aspects of endoscopic retrograde cholangiopancreatography (ERCP) requires several years of practical training and continuous refinement of knowledge and skills. ⁸ With the advent of noninvasive tests such as magnetic resonance cholangiopancreatography (MRCP) and endoscopic ultrasonography (EUS), ERCP is an almost purely therapeutic procedure. This is creating a new challenge in the training of young endoscopists, as ERCP procedures are becoming more complex and are concentrated in large- or mid-volume endoscopy centers. ¹⁷ , ¹⁸

In most fellowship training programs, traditional ERCP training follows education in diagnostic gastroscopy and colonoscopy and is often begun when the trainee has been introduced to polypectomy, hemostasis, or EUS training as part of a “learning pyramid” (▶Fig. 1.3). ¹

Jowell et al ¹⁹ found that a minimum of 180 to 200 ERCPs are needed to be performed before a trainee could attain competency in ERCP. ¹⁹ (▶Fig. 1.4) Approximately, 80 to 100 ERCPs per endoscopist per year appear to be necessary to maintain adequate competence for biliary procedures and 250 ERCPs per endoscopist per year for complex therapeutic procedures in the pancreas. ²⁰ The ERCP volume plays a role in complication rates. In various studies, a minimum of 40 to 50 endoscopic sphincterotomies (ESTs) per endoscopist per year was found to be associated with a lower complication rate in comparison to endoscopists with a lower EST frequency. ⁸ , ²¹ Rabenstein et al ²² showed that both prior experience and ongoing volume of ERCPs influence the success and complication rate.

Now that most ERCPs are performed for therapeutic purposes, it is a matter of controversy whether cannulation is the next technique for the trainee to learn after he or she is able to maneuver the duodenoscope competently to the papilla. For example, it is well known that for routine stent exchanges in the setting of a prior sphincterotomy, fewer procedures (n = 60) are needed to obtain competence than is the case with cannulation of a native papilla (n = 180–200), and it is also known that stent exchanges are associated with a lower risk profile compared to cannulation. Patients with benign biliary strictures, chronic obstructive pancreatitis, and recurrent bile duct stones in the setting of prior sphincterotomy are also associated with lower risk during training.

The ASGE published their latest core curriculum for training in ERCP in 2016. ⁸ , ²³ Trainees who elect to perform ERCP should have completed at least 18 months of standard gastroenterology training, followed by at least 12 months of ERCP training.

Schutz and Abbott ²⁴ developed an ERCP grading scale based on procedural difficulty using benchmarks such as cannulation rates to gauge competency. A modification of this score was adopted by the ASGE as part of their quality-assessment document. Absolute numbers of procedures partially performed by a fellow may not realistically reflect competence. ²⁵ Where possible, trainee logbook records should specify particular skills completed by the fellow (cannulation, sphincterotomy, stent placement, tissue sampling), and should also indicate cases that the trainee completed without assistance. The ASGE guidelines state that most fellows require at least 180 ERCP cases before competency can be assessed, with at least half being therapeutic. ⁸ Although not all of the trainees may ultimately perform ERCP after the completion of their training, all fellows should at least develop an understanding of the diagnostic and therapeutic role of the procedure, including indications, contraindications, and possible complications. ²⁶

The decision by a program director as to whether to train one or more fellows each year to achieve sufficient competence will depend in some measure on the volume of ERCPs performed at the institution and the availability of experts in ERCP (▶Fig. 1.4). ¹⁹ For example, with an annual volume of 400 cases and three fellows, it would be reasonable to have one fellow perform 300 or more cases and provide the other two with an exposure to ERCP, rather than have all three individuals equally share cases, with a low likelihood that any of the three would reach competence by the end of the fellowship.