Defining competence

Competence in endoscopy may be defined as the ability to independently carry out procedures in a safe and effective manner, and across a spectrum of case difficulties and case contexts. For colonoscopy, this should cover the necessary periprocedural and postprocedural aspects according to national standards, set by the JAG,10 the BSG and the ACPGBI.3 The UK standards for colonoscopy published in 2016 contain guidance on the minimum key performance indicators (KPIs) required for competent colonoscopy.3 Guidance for tattoo placement and biopsies for chronic diarrhoea should be followed. On review by the working group, KPIs appropriate to reflect trainees’ performance summarised in table 2.

Terminal ileum (TI) intubation is indicated in cases of chronic diarrhoea, iron-deficiency anaemia, abnormal radiological imaging of the TI, right iliac fossa pain and suspected inflammatory bowel disease (IBD).11 12 The manoeuvre is associated with longer procedural times, and when performed in all cases, the benefit on diagnostic yield is unclear.13 In cases of suspected IBD, performing terminal ileal intubation as a part of colonoscopy is recommended by both European Crohn’s and Colitis Organisation and BSG IBD guidance.14 15 Terminal ileal intubation rates of 80%–85% are feasible in routine practice, although>600 lifetime procedures are required to achieve rates >85%.16 In order to set a minimum KPI for trainees, the working group proposed a minimum unassisted terminal ileal intubation rate of 60% when indicated for the procedure.

An independent endoscopist should be able to recognise and differentiate between normal and abnormal findings, take appropriate action during the procedure and write a comprehensive report. These cognitive skills are necessary for effective patient management, documentation and handover. Trainees can increase their experience with increasing exposure and case mix during their training lists.

ENTS are complementary to technical skills and are important for a safe and effective procedure.17 They involve cognitive, interpersonal and social skills and primarily consist of communication and teamwork, situational awareness, leadership, judgement and decision-making.18 ENTS are essential component of practice with potential positive effects on team performance and clinical outcomes.19 ENTS training, when delivered to an experienced endoscopy team, demonstrates significant improvements in knowledge and attitudes towards patient safety.17 In a recent randomised controlled trial (RCT), novice trainees exposed to ENTS training were rated to be more competent during their first hands-on colonoscopy procedures.20 All JAG direct observation of procedural skills (DOPS) and direct observation of polypectomy skills (DOPyS) assessment forms were updated in July 2016 to emphasise and objectively measure ENTS competencies.21

Although colonoscopy is a relatively safe procedure and severe adverse events are rare, they can be potentially life threatening with a mortality rate of 0.007%–0.07%.22 Complications can range from minor (eg, vasovagal reaction), moderate (eg, cardiac arrhythmias, side effects of sedation) or severe (eg, bleeding and perforation). They should be measured over the 30-day period after the procedure and may be classified as intraprocedural, postprocedural (<14 days post) or late (>14 days).23 Larger studies report perforation rates of 0.005%–0.085% and post colonoscopy bleeding at 0.001%–0.687% (depending on the indication), with an overall rate of 0.05% (1 in 2000)24 25 A competent endoscopist should be able to recognise complications promptly, apply appropriate endoscopic and non-technical skills and initiate appropriate management. In particular competency would include demonstration of immediate leadership of the endoscopy team in managing the complication, appropriate communication with other teams if needed and duty of candour in discussions with patients and their next of kin.

The quality of a colonoscopy procedure can be adversely affected by the quality of bowel preparation, patient comfort, technical challenges and other intraprocedural factors (eg, complications). A competent endoscopist should be able to recognise the limitations of the procedure and exercise judgement and decision-making to recommend or arrange subsequent investigations to explain the patient’s symptoms (ie, if necessary with a repeat procedure or alternative investigation).

A recent systematic literature review of polypectomy learning curves showed that polypectomy competence was achieved after completion of 250–400 polypectomies and after 300 colonoscopies.26 In addition, the rate of competency in polypectomy did not correlate with the established colonoscopy quality metrics (ADR and withdrawal time).27 Setting a mandatory minimum number of polypectomy procedures would not be achievable within the current shape of training and therefore competency should be driven by performance in DOPyS assessments as validated comprehensive polypectomy assessment tools.

The SMSA scoring system comprises four factors which determine the complexity of a polypectomy; (S) Size, (M) Morphology, (S) Site and (A) Access.28 Application of SMSA divides complexity of polypectomy into four levels: level 1 (4–5), level 2 (6–8), level 3 (9–12), and level 4 (>12). Defining the difficulty level of polypectomy aids therapeutic decision-making, correlates with adverse events and can avoid unnecessary repeat procedures.29 30 For larger polyps, the SMSA level can determine the appropriate colonoscopist and time slot allocation for a future polypectomy.

Polyp size is an important predictor of polypectomy difficulty, postprocedural outcome and surveillance intervals.28 However, estimates of polyp size during colonoscopy often exceed those measured from histology samples. Anderson et al compared endoscopic and histological size from 1528 polyps. Of all 222 polyps estimated as ≥1 cm on endoscopy, 46% were <1 cm on pathology, while of 1306 polyps estimated as <1 cm, 3.9% were ≥1 cm on pathology.31 Eichenseer et al found that in 63% of their polyp cohort, a size difference of at least 33% was detected between in situ and postfixation measurements, leading to inappropriate surveillance recommendations in over a third of the cases, regardless of histology and the number of detected polyps.32 The use of an instrument (eg, open biopsy forceps) to calibrate size might increase accuracy in both experts and novices.33 34

In combination with size, polypectomy site (right vs left) can influence outcome, due to different anatomy characteristics and complication profiles. Different access scenarios (proximity to appendix, TI or diverticular segments, over a fold) could also make smaller lesions more challenging to resect. Additionally complete endoscopic resection is more challenging in lesions that have been resected previously or have advanced changes on optical diagnosis suggesting early cancerous change.

In vivo assessment of colorectal polyp morphology enables decision-making with regard to the appropriateness and mode of endoscopic therapy.28 The Paris classification provides standardised nomenclature for polyp morphology, classifying lesions to protruding (sessile, pedunculated, subpedunculated), flat (flat elevated or depressed) or depressed, although mixed morphology may be also present.35 Risk factors for submucosal invasion include Paris classification 0-IIa+c morphology, non-granular surface and Kudo pit pattern type V.36 The Paris classification can be predictive of submucosal invasion, with rates ranging from 1.4% for the most commonly observed lesions (Paris IIa granular), 7.5% (Paris 1s) to 31% for Paris IIc or IIa+IIc lesions.36 Paris classification aids decision-making with polypectomy and should be used for the documentation of polyp morphology and be included in the endoscopy report.

Advancement of imaging techniques, new classification systems and the use of artificial intelligence have allowed for increased viability of optical diagnosis.37 To date, many classification systems have emerged (eg, Sano, Narrow-band imaging International Colorectal Endoscopic Classification (NICE), Japan NBI Expert Team (JNET), Simplified Identification Method for Polyp Labelling During Endoscopy (SIMPLE), Kudo, Workgroup serrAted polypS and Polyposis (WASP), BLI Adenoma Serrated International Classification (BASIC)) with some more widely used than others. Competency in validated optical diagnosis systems is critical in assessing the realtime malignant potential of polyps.

Kudo et al first described five different ‘pit patterns’ according to mucosal surface seen by magnifying endoscopy, as a helpful tool to predict histology.38 Specific analysis of lesions with Kudo type V pit pattern found a vastly higher incidence of malignancy than with other pit pattern types (56% vs 4.4% (pit pattern III) vs 5% (pit pattern IV) vs 0% (pit patterns I+II), n=479, p<0.001).38 The Sano classification was described in 2006 using NBI and was based on vascular patterns.39 In an attempt to simplify the process and make use of a more universal system, the NICE classification was created in 2009 and was based in colour, vessel and surface pattern, and has been widely implemented.40

Although using the validated NICE classification for real-time prediction of polyp histology for small lesions (<10 mm) was not superior to high definition white light endoscopy in a German RCT,41 data appraising the ‘resect and discard’ strategy showed that the use of NICE classification might be able to prevent leaving in situ high-grade adenomas or small invasive colorectal cancers42 and a further Japanese study supports NICE being a valid tool for predicting deep submucosal invasive carcinomas (SMIC) with overall sensitivity and negative predictive value of 92%.43 One RCT found the modified Sano classification (MS) to outperform NICE for differentiating neoplastic polyps and predicting resectability.44

In order to address differences in surface patterns in elevated and superficial lesions, the JNET classification was developed and validated as a tool for hyperplastic polyps, sessile serrated lesions (SSL), adenomas and SMIC but was less applicable for shallow submucosal cancers.45 46 In a further attempt to create a classification inclusive of SSLs, the WASP classification was developed which led to a sustained increased in accurate diagnosis especially for SSLs.47

The BASIC is based on Blue Light Imaging and takes into account polyp morphology, crypt and vessel characteristics. It has high interobserver concordance and has been validated for diminutive colorectal polyps,48 with lesion recognition accuracy improving with training.49

The SMSA scoring system has been shown to predict critical outcomes of endoscopic mucosal resection (EMR), thereby validating its application for identifying high-risk lesions.29 In the Munich Polypectomy study,50 size and right side location were associated with major complications and specifically right sided lesions >10 mm and left sided lesions >20 mm. Lower levels of endoscopic clearance were also shown for more complex (SMSA level 4) compared with less complex lesions (SMSA levels 2 and 3) (87.5% vs 97.5%, p=0.009).51 Following Delphi group consensus, the SMSA system has replaced the previous system of assessing polyp complexity based on size alone (ie, where level 1: <10 mm; level 2: 10–20 mm), with competency in colonoscopy to include competency in at least SMSA level 2 polypectomy.

There is significant variation in the use of diathermy for polypectomy among endoscopists, with variable usage of current (blended or forced coagulation) by polyp morphology.52 53 Pure coagulation current is favoured by some due to its efficient haemostatic properties but delayed bleeding can occur and prolonged use might cause deep thermal injury. Blended current alters the current and blend delivered according to tissue resistance and was thought safer due to the rationale that it provides adequate cutting with effective haemostasis although immediate bleeding is observed.54 A recent RCT comparing the two diathermy modalities found no difference in risk of serious adverse events, complete resection rate or polyp recurrence.55 Pure cutting current should be avoided due to its high risk for immediate post polypectomy bleeding.56

Trainees should have a basic understanding of the electrosurgical unit and diathermy settings needed for safe polypectomy. A recent UK survey identified practice variations and knowledge gaps on the diathermy use which may not only affect safety in doing polypectomy but also quality of training delivered to subsequent generations of trainees.52 Understanding of diathermy should be supported through the Colonoscopy Basic Skills Course and teaching should also be delivered during polypectomy hands on simulation, enabling assessment and development of trainees’ skills in a protected and safe environment.57

Bleeding and perforation are the most common and serious complications of polypectomy. Endoscopists should be able to mitigate, recognise and manage such complications.58 Careful examination of the polypectomy resection site is important to check for deep mural injury (ideally with a recognised scoring system), bleeding or residual polyp.

For post polypectomy bleeding, haemostasis can be achieved in most cases with the use of through the scope (TTS) clips,59 direct thermal therapy with coagulation forceps or snare tip coagulation, or alternatively, coagulation therapy with argon plasma coagulation. There are no direct comparative trials between clips and thermal methods and they should be applied as required in an individual case-by-case basis, although caution should be taken to avoid prolonged thermal therapy in the resection site for risk of delayed perforation.60 Over the scope clips,61 haemostatic powders62 and self-assembling matrix forming gels63 may also be considered, with interventional radiology or surgical back up should endoscopic management fail.

For intraprocedural perforation, endoscopic closure is a safe and effective alternative to surgery and is thus the treatment of choice for select cases. In a systematic review of 24 cohort studies, successful closure was achieved in 90% of intraprocedural perforations using endoscopic methods.64

Although some skills might be acquired later with further experience and exposure, independent endoscopists should be competent with TTS clip placement and at least one other form of haemostasis. In addition to technical skills, demonstration of ENTS is also crucial to coordinate subsequent patient management.

Acquiring competence

The JAG accreditation standards for endoscopy services have set quality standards to ensure that all training centres deliver safe and effective training.65 Herein, each trainee should have a nominated trainer who performs to BSG standards, has received appropriate training as an endoscopy trainer (Train the Colonoscopy Trainer course) and is assessed regularly by trainees and peers (trainee feedback and through Direct Observation of Training Skills—DOTS). Training should be supplemented by access to recommended JAG courses, assessment and certification tools (JETS access) and appropriate supervision precertification and post certification. Finally, each certified endoscopy unit should maintain a suitable environment and have policies in place to support training opportunities in line with a trainee’s personal development plan.

Colonoscopy trainers should maintain their own competencies and safety standards as defined by the BSG.3 All trainers will have attended a JAG approved Train the Endoscopy Trainer for example, Train the Colonoscopy Trainer course before they are assigned to a trainee, with ongoing peer assessments to ensure these training skills are maintained as per current JAG standards.

The acquisition of higher ENTS competencies can lag behind the development of the technical skills of scope control and may be addressed through focused and structured ENTS training. In the RCT by Grover et al, structured simulation-based curriculum covering ENTS improved colonoscopy and integrative (ENTS) performance in novice trainees compared with self-directed simulator based training.66 More recently, an assessor-blinded RCT by the same group found that novice trainees exposed to a comprehensive non-technical skills curriculum (covering didactic teaching and case base scenario, virtual reality (VR) simulation, scenario training and specific non-technical skills feedback) had superior overall colonoscopy performance and non-technical skills ratings in their hands-on cases compared with the control group.67 Additionally, didactic teaching and multiprofessional in situ simulation training can be used to improve ENTS.17 68 69

The Basic Skills in Colonoscopy course was implemented by JAG in response to poor performance demonstrated in the 2004 UK wide colonoscopy audit.45 The main objective of a JAG approved training course is to standardise the delivery of endoscopy theory and hands-on techniques to all trainees. This allows trainees to build on the knowledge acquired and improve their skills until certification.

A 1-week hands-on colonoscopy course showed positive, sustained improvement on key areas of skills acquisition.70 An interrupted time series analysis found that attendance of the Basic Skills in Colonoscopy course results in a step-change improvement in performance when measured using the composite PICI endpoint in all trainees regardless of procedural experience.71 However, the greatest benefit was found in trainees at earlier stages training (in the <70 lifetime procedure count group). We therefore advise that the course should be undertaken early in the individual’s training journey and preferably within their first 70 colonoscopies.

The use of simulators for diagnostic colonoscopy training has been well studied, as in theory it reduces risk exposure to patients and allows development of basic techniques. The current literature includes 5 meta-analyses and a total of 10 RCTs for colonoscopy, although the quality of evidence is limited by the risk of bias.72–76 Overall, simulators show good validity, can distinguish between competency levels and can improve endoscopic procedure completion and mucosal visualisation compared with those who receive no training.77 Within individual RCTs, VR simulation training has been shown to improve CIR in the first 10 colonoscopies,78 increase competency during the first 100 cases with higher completion rates.79 However, a meta-analysis of seven studies found no improvement in patient-based CIR in trainees exposed to VR training.76 While VR simulators are effective as a precursor to patient-based training, there is insufficient evidence for VR to replace early hands-on training or to count towards a trainee’s lifetime procedure count.

The ASGE Preservation and Incorporation of Valuable Endoscopic Innovations committee felt that the threshold for benefit to recommend widespread simulator use would be a >25% reduction in median number of clinical cases to attain competency.80 This has not been tested in an RCT.

Regarding polypectomy, the optimal simulator approach has not yet been established. Interim results of a randomised trainee cohort study have failed to show a significant increase in DOPyS performance in the group that received specific augmented reality training in addition to conventional simulator-based training.81 At present, it is unclear what proportion of polypectomy training should consist of cognitive didactics, use of simulation/ex vivo models, observation of experts or supervised hands-on polypectomy. There is a lack of robust data on the efficacy and effectiveness of training interventions which increase trainees’ competencies and this has led to practice variation in different institutions and countries.82 In a study of eight gastroenterology trainees who underwent a lecture-based training which included basic principles of polyp characterisation, polypectomy technique, outcome and management of complications, the training did not result in improvement in overall competencies assessed by DOPyS.83 Different training methods such as educational videos are increasingly used and have demonstrated effect on the learning curve of polypectomy skills.84 Studies supporting the use of simulators for polypectomy are limited. The Welsh Institute for Minimal Access Therapy colonoscopy suitcase is an ex vivo porcine simulator for polypectomy that has shown content validity for training in polypectomy skills. The simulator was validated for snare polypectomy and correlated with the real-life level of expertise of the user.85

Training with GI simulators can increase early clinical performance but the optimal manner to deliver training is still under review. Feedback appears to be essential to derive benefit from simulation training.66 86 87 As such, simulation-based training should be directly supported with coaching and feedback from appropriately skilled trainers to maximise efficacy.88 89

Only 50% of UK trainees achieve competency in polypectomy of lesions between 1 and 2 cm within 3 years of their first DOPyS.90 Additionally, more recent data suggest that the polypectomy learning curve requires 250–400 polypectomies or 300 procedures to achieve competency.26 Colonoscopy trainees should be competent in the resection of SMSA level 2 polyps. Thus to allow sufficient time for that and shorten the learning curve, training in polypectomy should start early in colonoscopy training, once basic handling and tip control are mastered, and continue in parallel to ongoing diagnostic colonoscopy skills training

Endoscopic training in therapeutics should start when the trainee has developed some early essential skills (basic tip control) and continue throughout their training. Procedures can harbour life-threatening complications or be part of a high-risk clinical scenario, and prove challenging even in experienced hands.

One way to increase confidence of the trainee and to make the task safer is to deconstruct and learn the different steps of a new technique in a more controlled, less stressful environment of a hands on model course with expert trainer support and clear learning objectives appropriate to endoscopy training. The existing JAG accredited model courses, Hands-on Polypectomy Skills Course and the pilot Upper Gastrointestinal Haemostasis course apply these principles.

Supportive evidence for improvement in task competency is lacking although one RCT reported significant improvements in the performance of polypectomy, control of upper GI bleeding and oesophageal dilatation following a hands-on course.91 Additionally, a recent evaluation study of the 1 day BSG-JAG developed pilot haemostasis course in two centres showed that trainee confidence increased in both in their understanding and delivery of main therapeutic techniques in the management of upper gastrointestinal bleeding.92

As SMSA level 2 competency is the minimum requirement for competency in colonoscopy, trainees should be familiar with all the different techniques that might be used in endoscopy to safely and effectively remove all such polyps. Cold biopsy polypectomy (CBP) is a technique used for diminutive polyps (≤5 mm) but has been associated with high rates of incomplete resection, with only 39% of polyps being completely resected using CBP.93 The use of jumbo biopsy forceps is superior to standard CBP for small, sessile polyps although the efficacy remains questionable (78.8% vs 50.7%).94 When CBP was compared with cold snare polypectomy (CSP), the rate of complete resection was significantly higher in the CSP group for diminutive and small colorectal polyps, although for those polyps <4 mm complete resection rates were similar.95 96

CSP is favoured for small sessile lesions due to its safety profile, speed of resection and effectiveness. The CRESCENT study (A comparison of the resection rate for cold and hot snare polypectomy for 4-9mm colorectal polyps: a multicentre randomised controlled trial) and a recent meta-analysis comparing CSP with hot snare polypectomy (HSP) for polyps <10 mm reported similar complete resection rates of 92%–98%, but lower delayed bleeding rates and shorter procedure times for the CSP groups.97–99 CSP is unable to achieve en-bloc resection for lesions >10 mm,100 101 although lower residual polyp rates are seen with SSLs compared with adenomas (1.1% vs 11%) indicating a selective role for piecemeal CSP.102 103

HSP is the preferred technique for pedunculated polyps but also for larger lesions (>10 mm) and especially adenomas as part of lift-assisted polypectomy (EMR). It is associated with higher risk of perforation and also delayed bleeding depending on the current/setting used and type of polyp removed.

Trainees should be aware of the strengths and weaknesses of each method and be trained to apply these appropriately.

Trainees should receive a structured training approach to gain competency in use of the Paris polyp classification and at least one validated optical diagnosis system to classify and document polyps (statements 1.9 and 1.10). These could include NBI International Colorectal Endoscopic (NICE),41–43 104 JNET,45 104 BASIC48 105 and the WASP classification47 to improve diagnostic accuracy of hyperplastic and adenomatous histology and to predict serrated histology. Other proposed classifications such as ICE and SIMPLE for i-scan platform but they have not been fully validated in clinical practice yet.106 Ideally, training should consider encompassing multiple endoscopic platforms

Training modules can be didactic or web based and should take into account all commonly encountered lesions including hyperplastic polyps, SSLs, adenomas and cancer to allow effective implementation in clinical practice. Recently, Smith et al found no difference in diagnostic accuracy for the prediction of diminutive/small polyps between trainees assigned to didactic training and computer based training.107 The optimum mode of optical diagnosis training is therefore unclear.

Analysis of post colonoscopy colorectal cancers (PCCRC) highlight that 89% of these can be avoided with 8% of cases being attributed to detected polyps in the area of subsequent cancer not being removed. This would suggest inadequate lesion assessment (and subsequent decision not to proceed with resection) at the index procedure has an important role in development of these cancers.108

Optical diagnosis training should be incorporated into the Basic Skills in Colonoscopy course with a combination of precourse self-study material and in-course discussions to enhance understanding. Systematic, feedback-based training programmes have helped endoscopists with different levels of experience to develop high accuracy and good intraobserver agreement using NICE classification,109 while in-class teaching and self-directed learning using a standardised educational tool for these had similar results in accuracy of comparing adenomatous versus hyperplastic colonic polyps in a recent RCT.110

Polypectomy is a complex skill that requires significant training time. Lesion recognition and description with appropriate application of Paris classification and optical diagnosis platforms is an important first step and requires continuous practice, case variation and adequate exposure. Management planning and decision-making on optimal therapeutic approach for a lesion using SMSA score along with all other non-technical skills around polypectomy are also critical for a safe and effective resection technique. These discussions could precede hands-on technical skills experience and will aid future development. Regular feedback, use of DOPyS and reflective tools can facilitate these discussions.

The colonoscopy DOPS and DOPyS are formative assessment tools for diagnostic colonoscopy and polypectomy respectively which itemise specific competencies in sequential order.111 112 Each competency item is grouped within a domain, for example, preprocedure, procedural, management of findings, postprocedure and ENTS.111 All items covered within DOPS and DOPyS should be covered during colonoscopy and polypectomy training.

Water-assisted insertion techniques are now routinely used by many endoscopists around the world with notable benefits on patient comfort, increasing patient acceptance of repeat procedures, higher adenoma detection rates and procedural simplification.113–119 Colonoscopy trainees rated water-assisted colonoscopy (WAC) as easier to learn compared with air insufflation colonoscopy in a randomised crossover study involving three trainees.120 Lower rates of loop formation, as supported by magnetic endoscopic imaging (MEI) studies,121 may explain these benefits. WAC should be considered part of training in colonoscopy where expertise is available.

Scope tracking technologies such as MEI provides real-time three dimensional views of colonoscope configuration within the abdomen.122 MEI allows for direct visualisation of loop formation as well as the direction and degree of rotational manoeuvres required for directly observed withdrawal or precise external abdominal pressure.123 In training, use of MEI can facilitate discussion between the trainee and trainer to resolve issues with scope advancement or patient discomfort and explain decision making without taking over the scope.124 A meta-analysis of eight RCTs showed that MEI is beneficial in training and can improve performance with higher CIRs in inexperienced endoscopists and in experts in technically difficult cases.125 Some studies show improvement in caecal intubation times and patients comfort scores although data are conflicting.126 127 MEI can also improve the accuracy of preoperative localisation of colonic tumours, which is important when planning laparoscopic resection.128 129 Although MEI is a useful tool in colonoscopy and training, it is recognised that MEI is not universally available due to cost or compatibility with some endoscope manufacturers, and may be contraindicated in some patients. We therefore advise its use in training if available.

Although the learning curve to competency in colonoscopy has been studied, the findings are heterogenous and vary according to the endpoints by which competency is defined. Most studies apply an unassisted CIR of 90% as an indicator of competency in colonoscopy with no caecal intubation time limit. This was achieved after 233 procedures in the UK study by Ward et al 130 and in>90% of trainees after 500 procedures in the US study by Spier et al.131 The UK study by Siau et al based on DOPS assessments (N=1199 trainees) found that 250–299 procedures were required to attain competency in>90% of assessable skills within the procedural domain of DOPS, although skills such as ‘pace and progress’, ‘proactive problem solving’ required>300 procedures.111

When factoring in caecal intubation time limits, the time to competency is potentially longer. In a study of colorectal surgery trainees, >400 procedures was required to achieve this CIR>90% and caecal intubation time of≤10 min).132 In the study by Sedlack et al, the endpoint of CIR>90%, caecal intubation time<15 min and competent assessment scores (>3.5) was achieved in 90% of trainees after 300 procedures.133 When taking data from polypectomy learning curves into account, competency was achieved after 300 procedures.26

As it is likely that>300 colonoscopies will be required for competency in diagnostic colonoscopy and SMSA level 2 polypectomy, it is important to note that previously, in certain cohorts of trainees, undertaking this number of procedures has not been achievable during specialty training.134

Previously, JAG colonoscopy certification consisted of two stages—provisional and full. Based on JETS data, trainees apply for provisional certification after a median of 269 procedures (IQR 226–342) and for full certification after 403 (IQR 339–509).7 However, not all endoscopists applied for full certification once provisional certification was granted, and as such, competencies in higher level polypectomy could not be assured. Therefore, a simple single-stage colonoscopy certification process was preferable providing a uniform practice standard to patients and this has now been adopted with a minimum of 280 colonoscopies (based on Siau data)111 required before application.

JAG certification in flexible sigmoidoscopy is awarded when trainees meet competency standards evidenced in KPIs and DOPS,135 and are competent in the resection of small polyps (up to 10 mm). Some of the technical and ENTS skills in colonoscopy are generic to flexible sigmoidoscopy, although additional training is required for aspects such as sedation, loop management, terminal ileal intubation, postprocedure planning and the resection of lesions up to SMSA Level 2.

Development of earlier competence in colonoscopy in trainees in this cohort is supported by the finding that trainees with flexible sigmoidoscopy experience (>100 procedures) were more likely to achieve caecal intubation (OR 2.4, p<0.01).130 Additionally, a further study found that trainees who achieved splenic flexure intubation rates>90% also achieved the target of CIR>90% within a significantly lower number of procedures (mean of 208 vs 352 procedures, p=0.03).136 In trainees who have been awarded JAG flexible sigmoidoscopy certification, the working group recommended a minimum procedure threshold of 200 lifetime colonoscopy as an eligibility criterion to trigger summative assessment in colonoscopy.

Skills can decay over time. Jorgensen et al reported decrease in CIR after 6-week break for endoscopy trainees137 and this was confirmed in recent retrospective analysis looking at the impact of a 2-week break from colonoscopy practice on performance in independent endoscopists, which showed a small but significant drop in CIR and the composite score, performance indicator of colonic intubation.138

Current JAG training standards state that each trainee should have a minimum of 20 dedicated training lists a year in addition to access to ad-hoc training opportunities.65 With impending changes to structure of medical training, on-call rotations, dual accreditation commitments and the minimum procedure count of>280 procedures (statement 2.15) for certification, ensuring adequate training opportunity is important for trainees to achieve their certification during training. Increased availability and attendance of endoscopy training opportunities will be required to achieve this goal during specialty training.

DOPS assessments were developed to standardise skills assessment in a structured approach, to facilitate feedback provision and to support the certification process. They provide formal objective evidence of skills acquisition for both technical and non-technical skills.139 140 Accurate regular assessment of a trainees’ performance when learning a new skill can help them reach their full potential,141 while delivering constructive feedback after a procedure can stimulate reflection and and enhance learning.142 For training programmes, DOPS help to identify areas for improvement and when performed at regular intervals can be used to chart competency development. Indeed, the lifetime colonoscopy DOPS count is an independent predictor of procedural competency, while DOPS assessment scores strongly correlate with the CIR.111 The updates to DOPS forms in 2016, which incorporated ENTS and a change in scoring format from a performance-based scale to a supervision-based scale has also improved the validity of assessments.21

The updated requirements for certification include a minimum of 25 mandatory formative DOPS. While other training systems have recommended assessment blocks, for example, five consecutive assessments at intervals of 25 procedures, this was not found to affect competency development. For pragmatism, we advise a minimum of one DOPS per training list.

Similar to DOPS (statement 2.18), DOPyS should be performed regularly to chart competency development in polypectomy. Although DOPyS assessments should ideally be completed after every polypectomy performed to enhance performance, this may not always be feasible due to time constraints. We advise at least one DOPyS per training list where applicable.

Reflective practice is an important aspect of endoscopy and is encouraged in the General Medical Council standards for good practice. It empowers practitioners to develop self-awareness and insights on their strengths and weaknesses, optimises the learning process, consolidates practice and identifies opportunities for improvement.143 Trainees are encouraged to reflect on their clinical practice, endoscopic skills, interesting cases, ENTS, complications and so on, using the JETS ‘Reflection Tool’ at least once every 50 procedures. In line with GMC recommendations, this is one of the new requirements for endoscopy certification which should be reviewed by the endoscopy trainer during appraisal meetings.

Assessment of competence

The colonoscopy DOPS assesses 24 individual competencies under five broad domains (seven preprocedure; eight procedure; three management of findings; two post procedure; four ENTS).111 The current DOPS integrates a supervision-based scoring scale and an ENTS section which provides greater validity compared with the previous iteration.144

It is recognised that other competency assessment tools have been developed internationally which have undergone validation.77 140 145 The Assessment of Competency in Endoscopy (ACE) tool,146 which superseded the Mayo Colonoscopy Skills Assessment Tool, is currently used in North America.147 ACE measures 14 competency items and awards overall scores for technical and cognitive skills on a performance-based scale ranging from 1 (novice) to 4 (highly skilled). Following the analysis of 1061 ACE assessments for 93 fellows, an overall score of 3.5 was reported as the optimal competency threshold; this was attained for cognitive endpoints before technical skills, with competency achieved in nearly all items after 250 procedures. The Gastrointestinal Endoscopy Competency Assessment Tool,148 and (Skill Assessment in Fellow Endoscopy Training) SAFE-T,149 have also been introduced and supported by validity and reliability data, but not have been incorporated into national endoscopy training systems as with the DOPS. The Resident Practice Audit in Gastroenterology instrument assesses professionalism, interpretation of data, patient safety, knowledge, independence and technical skills, in addition to an overall score, and is being evaluated in Canada.150

As endoscopy trainees and trainers in the UK are familiar with DOPS, we recommend the continued use of DOPS as the default formative and summative assessment tool during colonoscopy training.

Endoscopy skills should be demonstrated across a breadth of different scenarios and with different case difficulties to ensure competency across a range of case contexts.140 DOPS assessments can be initiated by either the trainee or trainer but should be prospectively selected to minimise selection bias.

DOPS performance can be used to gauge when an individual is ready for summative assessment. Competent performance in colonoscopy DOPS should be demonstrated in more experienced trainees who have met or are approaching eligibility requirements for summative assessment. Prior to summative assessment, therefore, trainees should be competent in >90% of items assessed in DOPS,111 with up to 10% of items scoring ‘minimal supervision’ and no items rated as requiring ‘maximum supervision’ or ‘significant supervision’.

DOPyS is a validated polypectomy assessment tool for colonoscopy and flexible sigmoidoscopy112 which has been endorsed by the US Multi-Society Task Force on colorectal cancer.151 The use of DOPyS improves trainees documented exposure to therapeutic endoscopy, provides formal evidence of polypectomy skills acquisition and serves as an effective tool for assessing and certifying polypectomy in the UK.139 DOPyS was also used by Patel et al in the validation of Cold Snare Polypectomy Assessment Tool (CSPAT) which was developed specifically for CSP.152

Statement 2.9 highlighted the importance of applying the correct method for polypectomy according to the size and type of the polyp to reduce risk and improve outcomes. These techniques might be required to remove an SMSA level 1 polyp and thus trainees need to demonstrate competence in each modality. A systematic review on the learning curve for polypectomy based on ‘independent en bloc resection’ and delayed post polypectomy bleeding rate suggests that 250–400 polypectomies are required.26 The I-DOPyS study suggested a median of 15 polypectomies to achieve competence in CSP.153

We recommend a minimum of two assessments for SMSA level 1 polyps at the standard of ‘competent for independent practice’ for each modality (CSP, diathermy assisted resection of stalked polyps and diathermy assisted EMR). In total, this constitutes a minimum of six competent DOPyS for SMSA Level 1 polyps with the four most recent SMSA Level 1 DOPyS scoring ‘competent for independent practice’ in all items.

For independent practice in colonoscopy, competency in SMSA Level 2 polypectomy is required. This replaces the previous competency requirement of being expected to resect polyps up to 20 mm in size. Studies on the polypectomy learning curve recommend that trainees perform 50 EMRs with mentorship to prevent unacceptably high polyp recurrence rates.26

As per statement 3.5, trainees will need to have achieved competence in SMSA Level 1 polypectomy. In addition to this, trainees should complete a minimum of two DOPyS for each polypectomy technique (CSP, diathermy-assisted resection of stalked polyps and diathermy assisted EMR), all performed for SMSA Level 2 polyps. This constitutes a minimum of six competent DOPyS for SMSA Level 2 polyps with the four most recent SMSA Level 2 DOPyS scoring ‘competent for independent practice’ in all items.

Inadequacy of initial polyp resection contributes to PCCRC rates and analysis of these cases highlights that 7% of PCCRC cases may be attributed to this cause.108 154

In addition to DOPS, JAG certification adopts the use of minimum procedural numbers and KPIs from self-reporting methods to inform readiness for summative assessment. The KPIs for competent practice are aligned with trainee-relevant metrics featured within the UK colonoscopy quality standards (Statement 1.1). In line with previous JAG certification criteria,155 KPIs will continue to be measured over the last 3 months with a minimum of 15 procedures to enable sufficient practice volume to gauge performance. The eligibility criteria for summative assessment reflect the requirements described within this document.

To ensure objectivity of summative assessment, JAG had traditionally mandated DOPS to be scored by assessors who are not the trainee’s usual trainer (defined as the dedicated named trainer allocated to the trainee during their attachment who supervises the majority of training lists), involving a total of four summative DOPS performed within a 1 month window. This could involve different permutations, that is, as a 2+2 process simultaneously (2 assessors over 2 cases), as a 2+2 process sequentially (2 assessors over 4 cases), as a 2+1 + 1 process (3 assessors over 4 cases), as a 1+1 + 1 + 1 process (4 assessors over 4 cases), or as a 3+1 process sequentially (2 assessors over 3 cases).

Although there are data suggestive of the role for distant supervision, that is, through magnetic endoscopic imager configurations,156 video recordings,70 157 or live video transmissions on tablet devices,158 these are currently insufficient to be recommended as alternatives to directly observed assessments within the same room due to their limitations with ENTS assessments. Further research should be centred on these distant methods in conjunction with DOPS to facilitate external formative/summative assessments in colonoscopy.

Postcertification support

The transition between supervised training and newly independent practice can be a challenging period and may be the first opportunity to perform colonoscopy without an in-room trainer. Most newly certified practitioners will require a defined period of supervision (eg, 1 year) and ongoing endoscopy exposure to achieve the outcomes of an experienced practitioner.159 Assigning a named individual to meet regularly with a newly certified endoscopist to review progress and to discuss any clinical and non-clinical challenges faced during this period offers support and allows a relationship of trust and rapport to develop. Instances where such support may be beneficial may include: review of photodocumentation (and indeterminate lesions), postprocedural management plans, therapeutic decision-making. Such meetings should be conducted on a regular basis, to safeguard patient management and to facilitate skills development for the newly independent practitioner. The supervisor should possess valid credentials (appropriately experienced with competent KPIs) and have attended a Train-the-Trainers course.

The recent implementation of the UK National Endoscopy Database enables procedural data to be uploaded to a centralised server to benchmark an endoscopist’s KPIs against national quality standards.160 In accordance with JAG QA standards,65 performance data of all independent endoscopists should be reviewed at regular intervals to enable practitioners to discuss queries, concerns, raise issues with caseload or time management, and to receive support and further skills training where necessary. This should also include plans for support if underperformance is self-reported or observed during the newly independent period, in line with JAG guidance.161

As part of professional development, training requirements of newly certified endoscopists should be identified, discussed and recorded on an individualised personal development plan. For colonoscopy, this could include mentored lists (Statement 4.4) to improve specific technical or non-technical skills, or further skills training in therapeutics, or the attendance of multidisciplinary team meetings.

The path from competent to high-quality practice involves an ongoing learning curve post certification. Newly certified endoscopists need to be able to demonstrate all the appropriate technical and non-technical skills without a trainer in the room which can increase their mental workload and might lead to longer times to complete a test safely. Data from Siau et al showed that after colonoscopy certification, 18% of trainees had a dip in performance (CIR<90%) in their first 50 procedures, which decreased after 100 additional independent procedures.159 Endoscopy departments should allow for adjustments on the lists of the newly independent colonoscopists with reduced caseload if applicable, for example, for the first 6 months, to encourage familiarisation with independent practice and as confidence develops. It is advisable for progress to be reviewed by the named individual assigned to the newly certified endoscopist (Statement 4.1) and if applicable, to review caseload after the transition period.

Performing a minimum annual volume of colonoscopy procedures helps to maintain competency and is independently associated with improved patient outcomes including CIR,162 PICI and lower rates of adverse events.163–165 Harewood et al analysed performance of junior faculty members (<5 year experience) and reported significantly higher completion rates in those performing>200 colonoscopies per year than those who do not (92.5% vs 88.5%).162 The BSG guidance on KPIs recommend a minimum of 100 colonoscopies per year with an aspirational standard of 150 procedures per year to be undertaken by the endoscopist or directly supervising a trainee in the room.3

Within the Bowel Cancer Screening Programme framework on mentorship and QA, a mentor is defined as ‘a trusted counsellor or guide’. It is important for new certified endoscopists to feel supported by their department and be given access to mentored lists for a defined period, for example, first 6 months, after certification. A mentored list may involve a colleague with expertise either observing the newly independent endoscopist or vice versa. This can be useful to facilitate ongoing training, upskilling (eg, to SMSA 3+polyps), reflective practice and upskilling during this transition period, for supporting underperformance, or for imparting Train-the-Trainers skills to those who wish to eventually train others. Such an approach can also improve career satisfaction and mitigate burnout.166