We conducted a comprehensive literature search to identify studies related to pediatric endoscopy training and quality improvement. We searched electronic databases, including PubMed, Scopus, and Web of Science, for articles published between January 2000 and March 2023. The main reason was for this choice was the paucity of data before the year 2000. The search strategy included a combination of keywords and Medical Subject Headings (MeSH) terms related to pediatric endoscopy training, education, quality improvement, and patient outcomes. We used “pediatric”, “children”, “gastrointestinal endoscopy”, “training”, “competence”, “skill acquisition” “assessment”, “simulation” as keywords. The search was limited to studies published in English. Two reviewers independently screened the titles and abstracts of the identified studies for relevance to the review topic. The full text of potentially relevant studies was then reviewed in detail to determine their eligibility for inclusion in the review. Inclusion criteria for the studies were: (a) studies that reported on training strategies, education programs, or quality improvement initiatives in pediatric endoscopy; (b) studies that assessed the impact of these interventions on patient outcomes or provider performance; and (c) studies that were published in English. Exclusion criteria were: (a) studies that focused on adult endoscopy; studies reporting mixed pediatric/adult results were included; (b) studies that did not report on training or quality improvement interventions; and (c) studies that were not published in English.

Two reviewers independently extracted data from the included studies, including study design, study population, intervention type, outcomes assessed, and results. Any discrepancies in data extraction were resolved by consensus or by a third reviewer if necessary. The data were synthesized narratively, with a focus on identifying common themes and trends in the literature related to pediatric endoscopy training and quality improvement.

The quality of the included studies was assessed using the Cochrane Risk of Bias tool for randomized controlled trials (9) and the ROBINS-I tool for non-randomized studies (10). Two reviewers independently assessed the quality of the studies, and any discrepancies were resolved through discussion or by a third reviewer if necessary.

As this study is a narrative review, no ethical approval was required. This narrative review was reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement for reporting systematic reviews and meta-analyses (11).

Typically, training in GI endoscopy is conducted in a “hands-on” setting, with a trainee learning to perform the procedure under the direct supervision of an experienced endoscopist. This model surely has some clear benefits since the one-to-one relationship with an experienced trainer provides the opportunity of immediate feedbacks during training. However, there are some obvious shortcomings in taking the first steps in GI endoscopy in a clinical setting and on actual patients. Indeed, it can be difficult to embrace an overload of information and feedbacks in a stressful environment, where the “trial and error” can increase discomfort and the risk of complications for the patients. In addition, time is a tyrant, and clinical demands can limit the trainer's ability to provide comprehensive directions and feedback during procedures. This issue is particularly relevant in the pediatric setting, as parents and trainers tend to be more protective of children, resulting in fewer cases available for trainees to work with (19). To address this challenge, integrating instructional aids into GI endoscopy training curricula could be a potential solution (20). Such an approach can provide a safe learning environment for trainees to develop their skills without harming patients, especially in the early stages of their learning. Simulation-based training is an effective way to achieve this goal, as it allows trainees to practice essential skills repeatedly, receive feedback from experts, and gradually improve their abilities (21). Over the last few decades, endoscopy simulators have been developed and improved continuously. These simulators can be broadly classified into four main types: mechanical models, in vivo and ex vivo animal models, and virtual reality (VR) simulators (22). While mechanical models were the first to be developed, live animal models are considered more realistic, but they come with significant drawbacks such as high costs, specialized facilities, deterioration over time, and ethical issues (23). To overcome some of these downsides, ex vivo models have been developed, combining animal organs with plastic parts. VR simulators, combining visual and haptic interfaces, present a wide range of scenarios resembling reality and/or tasks that are planned to train the learners in specific skills (24).

Simulation-based learning programs have proven effective in supporting the classic apprenticeship model. A systematic review with meta-analysis of 21 randomized trials enrolling 1,181 participants by Singh et al. showed that simulation-based training in GI endoscopy significantly improved process skills and behaviors both in a test setting and in clinical practice. Additionally, it improved time to procedure completion and patient outcomes, such as procedural completion rate and risk of major complications (25). Another recent systematic review with meta-analysis focused on VR endoscopy simulation, included 18 randomized controlled trials with 421 participants and 3,817 endoscopic procedures. The study revealed that simulation-based endoscopic training was beneficial in complementing the performance of novice endoscopists before patient-based training, offering some competence-related benefits such as the rate of independent procedure completion, overall training performance and mucosa visualization (26).

There is limited research on the use of simulation-based training in pediatric endoscopy. Computer-based simulation training has been shown to improve trainees' confidence and technical skills, as measured by self-report assessments (27). However, a recent survey conducted among 71 pediatric GI fellowship program directors in North America and Canada highlighted that only slightly more than half of the programs integrate simulation-based training in their curricula, despite recognizing its importance, particularly for novices in GI endoscopy. The most commonly perceived barriers to its wider implementation in training programs were the equipment costs and the lack of a validated pediatric simulation curriculum (28). Providing the trainees with a simulation-based technology does not guarantee effective use. Trainers need to choose how to apply simulation-based learning to maximize its benefits for the novice endoscopists. Feedbacks is a crucial factor in procedural improvement, both in clinical and in a simulation settings (29). In the simulated setting, trainers can provide feedback at the end of task completion (terminal feedbacks) without concerns for patient safety, whereas this strategy is rarely feasible in the clinical setting. Walsh et al. found that the timing of feedback is a determinant of skill acquisition in the simulated setting (30). The authors reported that terminal feedback were more effective than feedback provided during task execution (30), as constant feedbacks may divert the trainee's focus from their performance to the feedback itself. Recent studies have focused on developing structured curricula for simulation-based endoscopy training. One such curriculum was designed by Grover and colleagues, which was divided into six hours of lectures and eighth hours of virtual simulator-based training with an expert tutor support. In a single-blinded randomized controlled trial, novice endoscopists assigned to the standardized curriculum demonstrated an improvement in technical, cognitive, and integrative skill achievement. They also exhibited an enhanced skill transfer into clinical setting, compared to those in a self-regulated learning group (31). Another study conducted a randomized controlled trial that included 39 novice endoscopists, who were assigned to either a control curriculum or a novel simulation-based curriculum that included dedicated non-technical skill (NTS) training. The NTS group outperformed the control group during two clinical colonoscopies, on the VR simulator, and in the integrated scenarios. The authors concluded that incorporating NTS training in simulation-based curricula is crucial for enhancing the overall performance of novice endoscopists (32).

To optimize the benefits of simulation-based training for pediatric GI endoscopy, it is essential to have a structured curriculum that appropriately challenges trainees. Grover and colleagues reported that a progressive simulation-based curriculum resulted in superior technical and non-technical skills, as well as an improved transfer to clinical setting (33). The findings emphasize the importance of matching the difficulty of tasks with trainee's ability to enhance the learning process. Additionally, less-technologically-advanced and less-expensive simulator may be optimal for basic skills training, reducing the financial burden of implementing simulation-based training in GI endoscopy.

Simulation-based training has proven beneficial, especially in the earliest phase of training. It helps accelerate learning curves by allowing novice endoscopists to become familiar with the endoscopic equipment, practice skills deliberately, and receive prompt, comprehensive feedback from an expert trainer without any harm to the patient (34). These features are particularly advantageous for pediatric GI endoscopy training, where even tertiary centers may have lower patient volumes, and there are greater concerns about taking the first steps in endoscopy with pediatric patients. However, efforts are still needed to enrich the literature with high-quality, specific pediatric evidence and to develop a standardized curriculum focused on the unique aspects of pediatric GI endoscopy.

Historically, competency in gastroenterology has been measured by the number of procedures a trainee has performed. However, these thresholds have generally based on expert opinion and may not accurately reflect a trainee's operative ability (39, 40). Furthermore, logbooks that record clinical experience have also been criticized for their reliance on procedural volumes, which can vary significantly among trainees and do not necessarily indicate technical competence (41). To address these issues, some guidelines have proposed the concept of a “competence threshold”, which is the number of supervised procedures required before a trainee's technical competence can be reliably assessed (4). Cass et al. reported an 80% success rate in esophageal intubation after completing 100 procedures (42), while another study showed 140 colonoscopies were required to reach a 90% rate of cecal intubation (43). However, more recent studies have suggested that these numbers may be higher, with some estimating that over 300 colonoscopies are required to reach a 90% technical success rate (44, 45). Establishing competence threshold for pediatric GI is particularly challenging, given the smaller number of endoscopies performed in children and the lack of high-quality literature on learning curves for pediatric endoscopists. Discrepancies have been observed among the guidelines that establish the competence threshold for both pediatric GI endoscopy since they have been largely extrapolated data from adult studies (19). Table 1 summarizes the competence thresholds as defined by different adult and pediatric GI scientific societies.

Additionally, obtaining the recommended number of specialized procedures required to meet competency standards may be difficult, even in tertiary facilities (53, 54). In a survey of 50 third-year pediatric GI fellows, only slightly more than half had performed more than 100 colonoscopies (55). Despite these challenges, it is crucial to establish competence thresholds to optimize learning and ensure patient safety. However, more research is needed to determine appropriate thresholds for pediatric GI endoscopy, as well as to develop reliable and valid assessment tools that can accurately measure technical competence.

Broekaert et al. recently conducted a survey conducted among young members of the European Society for Pediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) to evaluate the quality of pediatric endoscopy training across European centers. Despite completing a median of 200 upper GI endoscopies and 75 colonoscopies (both above the threshold suggested by the ESPGHAN syllabus), only 43% of the 68 surveyed trainees achieved a terminal ileum intubation rate >90%, highlighting the need of a more rigorous and standardized assessment of competency (56). The recent focus on optimizing quality in endoscopy has influenced the assessment of competency, with many scientific societies incorporating quality metrics in their credentialing guidelines (Table 1). However, while quality metrics may be a useful indicator of trainee's overall performance, they do not provide continuous feedback or outline criticisms or deficiencies. Adenoma detection rate applies poorly to pediatric GI endoscopy, since colonoscopies in children are infrequently performed with cancer screening purposes. In contrast, terminal ileum intubation rate is more important when in pediatric colonoscopy, as colonoscopy is often performed to diagnose inflammatory bowel disease, which requires complete ileal intubation, exploration, and sampling (57). The recently published guidelines from the Pediatric Endoscopy Quality Improvement Network (PENQUIN) working group set minimum targets for defining high quality ileocolonoscopy as an unadjusted cecal intubation rate of ≥90% and an unadjusted terminal ileum intubation rate of ≥85% (3, 8).

The completion of a certain number of supervised procedures does not guarantee that a trainee is adequately prepared to perform GI endoscopy independently. Rather, these numbers can be used as an adjunct to the more focused formative assessment. To document progress and proficiency level during endoscopy training, the recently published joint NASPGHAN/ESPGHAN guidelines from the Pediatric Endoscopy Quality Improvement Network working group recommend the use of competence assessment tools with strong validity evidence (8). However, there is currently limited literature on whether the implementation of these tools can improve clinical outcomes. Despite this, the Pediatric Endoscopy Quality Improvement Network working group has reached a consensus on the routine use of certain well-validated tools for the assessment of competence during GI endoscopy training. One of these tools is the direct observation of the procedural skills (DOPS) which is well-established among adult practitioners in the UK. DOPS assessments are performed by a trainer observing a trainee and are submitted electronically to the Joint Advisory Group on Gastrointestinal Endoscopy Electronic Training System e- Portfolio (37, 52, 58). Recent studies have provided low-quality evidence to support the use of DOPS scores to describe both technical and non-technical competency during pediatric upper GI endoscopy and ileocolonoscopy with sufficient sensitivity and specificity (59, 60). However, the only tool specifically developed for the assessment of pediatric ileocolonoscopy is the Gastrointestinal Endoscopy Competency Assessment Tool for pediatric colonoscopy (GiECAT_KIDS) (61). This tool was developed through a Delphi process by 41 North American pediatric endoscopy experts who established key colonoscopy aspects for proficiency acquisition. After item reduction and gradation, 18 checklist and 7 global rating items were generated, reflective of technical, cognitive, and integrative competency required for safe and proficient endoscopy (61). The same group validated GiECATKIDS by assessing 104 colonoscopies from 56 endoscopists across 3 North American centers, demonstrating the tool's strong reliability and validity as a measure of pediatric colonoscopy performance (62). Despite this rigorous development and validation process, the tool is not widely utilized.