Real-world practices and barriers in endoscopic submucosal dissection training: International comprehensive survey

Abstract

Background and study aims

Endoscopic submucosal dissection (ESD) enables curative resection of superficial gastrointestinal neoplasia but requires structured training. Although European Society of Gastrointestinal Endoscopy (ESGE)/American Society for Gastrointestinal Endoscopy (ASGE) curricula exist, real-world practices and barriers remain unclear.

Methods

We conducted an anonymous web-based survey (April–July 2025) with 55 items assessing training opportunities, barriers, curriculum adherence, and quality metrics, stratified by role and annual ESD volume.

Results

We received 288 responses (137 trainees, 151 trainers) from 39 countries across five geographic macro-areas. Nearly half (47.2%) had undertaken structured training, but only 46.5% fulfilled ESGE/ASGE curricula, with marked variation by center volume (69.6% in high-volume vs 26.8% in low-volume, P < 0.001). Lack of access to training models was reported by 46.5% overall, more frequently in low-volume centers. In vivo human training, animal models, and intensive courses were rated highest, whereas lack of simulators, fellowship costs, and limited time were leading barriers. Quality monitoring was reported in 62.5% of centers, prioritizing R0 and en bloc resection rates over procedure speed. Trainees valued traction and underwater techniques more than trainers. Regional differences were pronounced: North-West Europe and Asia-Pacific concentrated high-volume centers and greater curriculum adherence, whereas Africa/Middle East and South-East Europe faced resource and financial constraints.

Conclusions

This international survey reveals substantial disparities in ESD training and calls for coordinated efforts to improve access to structured curricula, mentorship, and affordable training models.

Keywords

Endoscopy Lower GI Tract - Polyps / adenomas / ... - Endoscopic resection (polypectomy, ESD, EMRc, ...) - Quality and logistical aspects - Training

Introduction

Endoscopic submucosal dissection (ESD) enables en bloc resection of superficial gastrointestinal neoplasms with high curative potential but demands advanced skills and prolonged training [1]. Japan is regarded as the benchmark for ESD training, offering trainees extensive exposure to high case volumes and a structured, stepwise supervision process. This approach enables them to achieve expert-level proficiency after roughly 80 supervised procedures [2]. In Western settings, the European Society of Gastrointestinal Endoscopy (ESGE) and American Society for Gastrointestinal Endoscopy (ASGE) have developed structured curricula for basic ESD skills acquisition, both emphasizing competency milestones, indirect quality metrics, and progression from ex vivo and in vivo models to independent practice, although differing in procedural thresholds and training structure [3] [4].

Despite these frameworks, real-world adherence to curriculum recommendations and access to key training resources such as simulators, animal models, and supervised patient cases remains uncertain, particularly outside high-volume expert centers [5]. Understanding global variability in ESD training conditions, as well as perceived barriers and unmet needs, is crucial to guide future educational strategies and promote equitable access to high-quality training.

This international survey aimed to characterize real-world ESD training across diverse geographic and institutional contexts, with a focus on curriculum adherence, available training pathways, and perceived barriers from the perspectives of trainees and trainers.

Methods

Survey design and inclusion criteria

We conducted a cross-sectional, anonymous, web-based survey across 39 countries worldwide.

A 55-item multiple-choice questionnaire was developed targeting training opportunities, perceived barriers, and curriculum adherence of ESD training.

Items included multiple-choice, open-ended, and Likert-scale questions on training opportunities, perceived barriers, and curriculum adherence. The Likert scale ranged from 1 (strongly disagree) to 5 (strongly agree), with 3 indicating a neutral response.

The working group that formulated the survey consisted of four members (GC, SF, MM, JJ). The questionnaire was developed by the working group based on available literature and expert consensus on ESD training. Prior to dissemination, it was informally pretested among the co-authors from different geographic and training backgrounds to refine wording and ensure face and content validity.

After approval by all components of the working group, the final version of the questionnaire was disseminated in English via the SoSciSurvey platform (https://www.soscisurvey.de) to trainers and trainees involved in ESD.

Survey invitations were circulated to endoscopists involved in ESD training through established professional and society-based mailing lists and networks. Distribution strategies were broadly comparable across geographic regions. Responses were collected at the individual level, with participation open to all eligible practitioners reached through these channels.

A first invitation was sent to all members, followed by two subsequent fortnightly reminders. The study took place from April to July 2025.

Definitions and objectives

For this study, “trainee” was defined as an endoscopist proficient in basic endoscopy and adverse event (AE) management who has started ESD skills acquisition under supervision. “Trainer” was defined as an experienced (at least 100 procedures performed) ESD endoscopist actively involved in training.

Institutional ESD volume was categorized as low (< 50 ESD/year), medium (50–100/year), or high (> 100/year), based on previous literature [6].

Geographic macro-areas were grouped as follows: Africa and Middle East, America, Asia-Pacific, Northwest Europe, and Southeast Europe based on an adapted version of the United Nations publication “Standard Country or Area Codes for Statistical Use” [7].

Within this framework, the survey was designed to address three main objectives: 1) to assess adherence to established ESGE/ASGE ESD training curricula, defined as self-reported fulfilment of key structural and training requirements; 2) to map available training pathways and educational resources across different geographic and institutional settings; and 2) to identify perceived barriers to ESD training and competency acquisition from the perspectives of both trainees and trainers.

Data collection and statistical analysis

Data were collected anonymously and exported into Microsoft Excel (Microsoft Corporation, Redmond, Washington, United States) for preprocessing. All survey items were mandatory, and respondents could not proceed to subsequent sections without completing each question; therefore, no missing data were present in the final dataset.

Descriptive statistics were used to summarize survey responses, with categorical variables presented as absolute numbers and percentages, and continuous variables as mean ± standard deviation (SD) or median with interquartile range (IQR), as appropriate. For Likert-scale items, results also were summarized as the proportion of respondents assigning a high relevance score (≥ 4), given clustering of responses at higher values and to improve interpretability.

Statistical comparisons were planned a priori to explore differences between trainees and trainers, and across institutional ESD volume categories and geographic macro-areas, in line with the predefined survey objectives. Comparisons between groups (trainees vs. trainers and across center volume categories) were performed using the Chi-square or Fisher’s exact test for categorical variables. When appropriate, multiple pairwise comparisons were performed for internal validation, applying Holm correction, but only the overall P value is reported. For parametric continuous variables, comparisons between two groups were performed using the independent samples t-test, applying Welch’s correction in cases of unequal variances as assessed by Levene’s test. For comparisons involving more than two groups, one-way ANOVA with Welch’s correction was used. For non-parametric continuous variables, the Mann–Whitney U test or Kruskal–Wallis test was applied as appropriate. A two-sided P < 0.05 was considered statistically significant. Statistical analyses were conducted using R Studio software (version 2025.05.0.496; R Foundation for Statistical Computing, Vienna, Austria).

Ethics approval

This study was conducted through an anonymous, voluntary online survey addressed to healthcare professionals. All subjects agreed to participate in the interview through informed consent for the collection, handling, and storage of data, which was included in the presentation of the questionnaire. In accordance with institutional regulations, ethics committee approval was not required for this type of survey.

Results

Participants and center characteristics

Of the 472 invited participants, 288 completed the survey, corresponding to a response rate of approximately 61.2%. A total of 288 respondents from 39 countries completed the survey, comprising 137 trainees (47.6%) and 151 trainers (52.4%). The proportion of males was 98 (71.5%) among trainees and 128 (84.8%) among trainers. Mean age was 36.2 ± 6.70 for trainees and 43.0 ± 7.30 for trainers.

Most participants were based in academic hospitals (62.8%), followed by tertiary referral hospitals (19.8%), secondary-level hospitals (7.6%), and first-level care centers (9.7%).

Geographic distribution covered all major world regions, with the highest representation from Southeast Europe (38.5%), followed by Northwest Europe (28.1%), Asia-Pacific (16.3%), America (8.7%), and Africa and the Middle East (8.3%).

Regarding institutional ESD workload, 42.7% of respondents were based in low-volume (LV) centers, 29.9% in medium-volume (MV) centers, and 27.4% in high-volume (HV) centers.

Nearly half the participants (47.2%) reported currently undertaking or having completed a structured training program, with significant differences between trainees and trainers (54.4% vs 40.4%; P = 0.001) and across different center volumes (LV = 36.6% vs MV = 60.5% vs HV = 49.4%; P < 0.001). Fulfilment of the ESGE or ASGE core curriculum was reported in 46.5% of cases, with significant variation across centers by volumes (LV = 26.8% vs MV = 53.5% vs HV = 69.6%; P < 0.001).

Most mentors (60.8%) had no formal pedagogical training, whereas only 18.4% had education specifically focused on ESD, with no significant differences based on ESD volumes (P = 0.104).

Regarding access to training models, 46.5% of participants reported no access (LV = 60.2% vs MV = 39.5% vs HV = 32.9%), 15.6% had free access (LV = 6.5% vs MV = 14.0% vs HV = 31.6%), and 37.8% had limited access (LV = 33.3% vs MV = 46.5% vs HV = 35.4%) (P < 0.001).

A detailed breakdown of participant demographics and center characteristics, stratified by role and annual ESD volume, is provided in [Table 1].

Table 1 Overall participant demographics and center characteristics, stratified by role and annual ESD volume.

Variable	Overall	Trainee	Trainer	P value	Low volume	Mid volume	High volume	P value
ESD, endoscopic submucosal dissection; ESGE, European Society of Gastrointestinal Endoscopy; IQR, interquartile range; SD, standard deviation.
Patients, n (%)	288	137 (47.6)	151 (52.4)		123 (42.7)	86 (29.9)	79 (27.4)
Mean age, y (SD)	39.8 (7.8)	36.2 (6.7)	43.0 (7.3)	< 0.001	40.9 (8.2)	39.3 (6.3)	38.5 (8.4)	0.07
Male gender, n (%)	226 (78.5)	98 (71.5)	128 (84.8)	0.010	93 (75.6)	71 (82.6)	62 (78.5)	0.49
Macro area, n (%)				< 0.001				< 0.001
Africa and Middle East	24 (8.3)	14 (10.2)	10 (6.6)		14 (11.4)	5 (5.8)	5 (6.3)
America	25 (8.7)	7 (5.1)	18 (11.9)		5 (4.1)	14 (16.3)	6 (7.6)
Asia-Pacific	47 (16.3)	11 (8.0)	36 (23.8)		25 (20.3)	7 (8.1)	15 (19.0)
Northwest Europe	81 (28.1)	47 (34.3)	34 (22.5)		19 (15.4)	28 (32.6)	34 (43.0)
Southeast Europe	111 (38.5)	58 (42.3)	53 (35.1)		60 (48.8)	32 (37.2)	19 (24.1)
Institution, n (%)				0.183				0.04
Primary care	28 (9.7)	18 (13.1)	10 (6.6)		16 (13.0)	7 (8.1)	5 (6.3)
Secondary level	22 (7.6)	12 (8.8)	10 (6.6)		15 (12.2)	3 (3.5)	4 (5.1)
Tertiary level	57 (19.8)	23 (16.8)	34 (22.5)		26 (21.1)	13 (15.1)	18 (22.8)
University Hospital	181 (62.8)	84 (61.3)	97 (64.2)		66 (53.7)	63 (73.3)	52 (65.8)
Median endoscopy exp, y [IQR]	10.00 [6.00–15.00]	7.00 [4.00–10.00]	13.00 [10.00–17.00]	< 0.001	10.00 [6.00–15.00]	10.00 [6.00–13.75]	10.00 [5.00–15.00]	0.72
Median ESD training, y [IQR]	4.00 [2.00–8.00]	2.00 [1.00–3.00]	7.00 [5.00–10.00]	< 0.001	3.00 [1.00–5.00]	5.00 [2.00–8.00]	6.00 [2.00–12.00]	< 0.001
Median ESD exp, y [IQR]	3.00 [1.00–7.00]	1.00 [0.00–2.00]	6.00 [3.50–9.50]	< 0.001	2.00 [1.00–4.00]	4.00 [2.00–7.00]	6.00 [2.00–10.00]	< 0.001
ESD exp, n (%)				< 0.001				< 0.001
0	36 (12.5)	32 (23.4)	4 (2.6)		22 (17.9)	6 (7.0)	8 (10.1)
1–10	40 (13.9)	31 (22.6)	9 (6.0)		32 (26.0)	2 (2.3)	6 (7.6)
11–50	67 (23.3)	44 (32.1)	23 (15.2)		38 (30.9)	20 (23.3)	9 (11.4)
50–100	50 (17.4)	27 (19.7)	23 (15.2)		19 (15.4)	18 (20.9)	13 (16.5)
100–250	35 (12.2)	0 (0.0)	35 23.2)		8 (6.5)	19 (22.1)	8 (10.1)
> 250	60 (20.8)	3 (2.2)	57 (37.7)		4 (3.3)	21 (24.4)	35 (44.3)
Selection for training, n (%)				0.126				< 0.001
Based on criteria	71 (24.7)	41 (29.9)	30 (19.9)		15 (12.2)	27 (31.4)	29 (36.7)
No formal selection	126 (43.8)	54 (39.4)	72 (47.7)		74 (60.2)	33 (38.4)	19 (24.1)
Personal recommendation or personal interest	91 (31.6)	42 (30.7)	49 (32.5)		34 (27.6)	26 (30.2)	31 (39.2)
Structured training, n (%)				0.001				0.004
No	116 (40.3)	40 (29.2)	76 (50.3)		62 (50.4)	28 (32.6)	26 (32.9)
Unsure	36 (12.5)	22 (16.1)	14 (9.3)		16 (13.0)	6 (7.0)	14 (17.7)
Yes	136 (47.2)	75 (54.7)	61 (40.4)		45 (36.6)	52 (60.5)	39 (49.4)
ESGE curriculum filling, n (%)	134 (46.5)	71 (51.8)	63 (41.7)	0.110	33 (26.8)	46 (53.5)	55 (69.6)	< 0.001

In a gender-stratified analysis, female respondents were younger than male respondents (37.1 ± 6.3 vs 40.5 ± 8.0 years; P = 0.002) and were less likely to be in a trainer role (37.1% vs 56.6%; P = 0.010). Accordingly, females reported shorter overall endoscopic experience (8.6 ± 5.3 vs 11.9 ± 7.3 years; P = 0.001), fewer years of ESD training (P < 0.001), and a lower cumulative ESD experience (P = 0.020).

Perceived impact of training paths and significant barriers

In vivo human training was the most highly prioritized training modality, being rated as highly relevant (score ≥ 4) by 87.5% of respondents, followed by in vivo animal models (79.9%), observership fellowships (76.0%), and hands-on simulator training (74.3%). Self-directed tutorials were also frequently rated as highly relevant (67.0%), whereas national and international hands-on fellowships received high ratings from 58.0% and 55.9% of participants, respectively. No significant differences were observed between trainers and trainees ([Fig. 1] a).

Fig. 1 Perceived impact of training paths, barriers, and training strategies according to role. a Different training modalities. b Training strategies. c Major barriers. d Quality indicators.

When stratified by role and center volume (Supplementary Fig. 1a), hands-on simulator training and observership fellowships were rated higher by participants from MV and HV centers compared with those from LV centers (P < 0.001 and P = 0.04, respectively).

Regarding perceived barriers, lack of access to simulator workshops emerged as the most relevant obstacle, being rated as highly limiting (score ≥ 4) by 56.6% of respondents. Lack of clinical cases (39.9%), high fellowship costs (42.4%), limited available time (40.3%), and limited access to mentors (41.0%) were less frequently rated as major barriers overall. Among these, only limited available time differed significantly between trainees and trainers, being more frequently reported by trainers (45.0% vs 35.0%; P = 0.018) ([Fig. 1] b).

In the stratified analysis (Supplementary Fig. 1b), high fellowship costs and limited available time for training were rated higher by LV trainers and trainees than by those from MV and HV centers (P = 0.002 and P < 0.001, respectively).

Gender-stratified analyses of training perceived barriers are provided in Supplementary Table 1.

Training modalities and preferred learning strategies

Participants most frequently rated as highly relevant (score ≥ 4) the following training steps: a progressive step-up approach from less to more challenging lesions (87.5%), starting with whole easy lesions (83.0%), dedicated ESU setting training (91.0%), and participation in a 1-week intensive course (78.1%). Trainees more often rated traction techniques (78.1% vs 64.9%; P = 0.002) and the underwater technique (56.2% vs 45.7%; P = 0.005) as highly relevant compared with trainers, whereas no significant differences were observed for pocket creation, closure-first strategies, or stepwise lesion difficulty ([Fig. 1] c).

Stratified analysis revealed clear patterns (Supplementary Fig. 1c): HV trainees more frequently rated traction (P = 0.04), underwater technique (P = 0.009), and the progressive step-up approach (P = 0.003) as highly relevant, while pocket creation was more often emphasized by LV trainers (P = 0.002).

Training modalities, quality metrics and personal considerations

Tracking of ESD quality indicators was reported in 62.5% of institutions, reflecting variable integration of quality monitoring into routine practice. The most frequently prioritized indicators (score ≥ 4) were R0 resection rate (89.6%), en bloc resection rate (88.5%), and overall AE rate (91.7%), followed by post-procedural perforation (84.7%) and intraprocedural bleeding control (62.5%). Dissection speed was less frequently prioritized (53.1%) ([Fig. 1] d). No specific pattern emerged in the stratified analysis (Supplementary Fig. 1d).

Maximum acceptable AE rates for trainees varied: 27.1% indicated < 5%, 21.9% < 10%, and 11.1% < 15%, whereas 22.9% reported no predefined threshold. Procedure time limits were supported by 34.4% of respondents, most commonly 60 minutes (IQR 40–60) for 10-mm antral lesions and 90 minutes (IQR 60–120) for 20-mm rectal lesions. The most frequently reported thresholds for independent practice were 10 to 25 cases (42.7%) and 26 to 50 cases (32.6%).

Training modalities, quality metrics and personal considerations

Networking strategies emerged as an important theme, with 75.0% of respondents rating the hub-and-spoke model as highly relevant and 72.9% supporting presence of at least one ESD-trained physician per center.

Financial aspects also influenced training accessibility: Willingness to invest in training varied significantly by role and center volume (P = 0.012). Trainees from HV centers more often reported self-funding €1000 to 2000 (50.0%) compared with those from LV (24.3%) or MV (14.3%) centers, whereas trainers from LV centers were more likely to rely on company sponsorship (17.1% vs 2.1% in HV centers). Institutional support was common across all groups, peaking among MV trainers (39.2%).

[Table 2] summarizes all ratings given by respondents to the abovementioned questions.

Table 2 Overall ratings of different survey items stratified by role.

	Overall (n = 288)		Trainee (n = 137)		Trainer (n = 151)		P value
	Median [IQR]	% rating ≥ 4	Median [IQR]	% rating ≥ 4	Median [IQR]	% rating ≥ 4
AE, adverse event; ESD, endoscopic submucosal dissection; ESU, electrosurgical unit; IQR, interquartile range.
Impact of training modalities, median rating [IQR]
Hands-on simulator training	4.00 [3.00–5.00]	74.3%	4.00 [3.00–5.00]	71.5%	4.00 [4.00–5.00]	76.8%	0.840
In vivo model training	4.00 [4.00–5.00]	79.9%	4.00 [4.00–5.00]	81.0%	4.00 [4.00–5.00]	78.8%	0.930
In vivo human training	5.00 [4.00–5.00]	87.5%	5.00 [4.00–5.00]	85.4%	5.00 [4.00–5.00]	89.4%	0.797
Observing fellowship	4.00 [4.00–5.00]	76.0%	4.00 [4.00–5.00]	76.6%	4.00 [4.00–5.00]	75.5%	0.329
Hands-on national fellowship	4.00 [3.00–5.00]	58.0%	4.00 [3.00–4.00]	58.4%	4.00 [3.00–5.00]	57.6%	0.577
Hands-on international fellowship	4.00 [3.00–5.00]	55.9%	4.00 [3.00–5.00]	52.6%	4.00 [3.00–5.00]	58.9%	0.196
Self-tutorial through videos	4.00 [3.00–5.00]	67.0%	4.00 [3.00–5.00]	67.9%	4.00 [3.00–5.00]	66.2%	0.951
Perceived barriers, median rating [IQR]
Lack of clinical cases	3.00 [2.00–4.00]	39.9%	3.00 [2.00–4.00]	39.4%	3.00 [2.00–4.00]	40.4%	0.578
High costs of fellowships	3.00 [2.00–4.00]	42.4%	3.00 [2.00–4.00]	38.0%	3.00 [2.00–4.00]	45.0%	0.163
Limited time for training	3.00 [2.00–4.00]	40.3%	3.00 [1.00–4.00]	35.0%	3.00 [2.00–4.00]	45.0%	0.018
Lack of mentors	3.00 [2.00–4.00]	41.0%	3.00 [2.00–4.00]	39.4%	3.00 [2.00–4.00]	42.4%	0.838
Lack of simulators	4.00 [3.00–5.00]	56.6%	4.00 [3.00–5.00]	57.7%	4.00 [3.00–5.00]	55.6%	0.861
Training modalities, median rating [IQR]
Pocket/conventional technique	4.00 [3.00–5.00]	59.0%	4.00 [3.00–5.00]	57.7%	4.00 [3.00–4.50]	60.3%	0.925
Use of traction	4.00 [3.00–5.00]	71.2%	5.00 [4.00–5.00]	78.1%	4.00 [3.00–5.00]	64.9%	0.002
Underwater technique	4.00 [3.00–4.25]	50.7%	4.00 [3.00–5.00]	56.2%	3.00 [3.00–4.00]	45.7%	0.005
Step-up by difficulty	5.00 [4.00–5.00]	87.5%	5.00 [4.00–5.00]	89.8%	5.00 [4.00–5.00]	85.4%	0.136
Start with whole easy lesions	5.00 [4.00–5.00]	83.0%	5.00 [4.00–5.00]	86.9%	5.00 [4.00–5.00]	79.5%	0.033
Start with defect closure	3.00 [2.00–4.00]	49.3%	4.00 [2.00–4.00]	51.8%	3.00 [3.00–4.00]	47.0%	0.616
Dedicated ESU setting training	5.00 [4.00–5.00]	91.0%	5.00 [4.00–5.00]	89.8%	5.00 [4.00–5.00]	92.1%	0.966
1-week intensive course	5.00 [4.00–5.00]	78.1%	5.00 [4.00–5.00]	79.6%	4.00 [4.00–5.00]	76.8%	0.090
Personal opinion on ESD network, median rating [IQR]
Hub-and-spoke model	4.00 [3.75–5.00]	75.0%	4.00 [4.00–5.00]	75.2%	4.00 [3.50–5.00]	74.8%	0.842
Access to easy cases in spoke	4.00 [3.00–5.00]	64.6%	4.00 [3.00–5.00]	67.9%	4.00 [3.00–4.00]	55.0%	0.019
≥ 1 physician per centre	4.00 [3.00–5.00]	72.9%	4.00 [4.00–5.00]	76.6%	4.00 [3.00–5.00]	69.5%	0.519
Quality indicators, median rating [IQR]
Overall AE	5.00 [4.00–5.00]	91.7%	5.00 [4.00–5.00]	94.2%	5.00 [4.00–5.00]	89.4%	0.832
R0 rate	5.00 [4.00–5.00]	89.6%	5.00 [5.00–5.00]	92.0%	5.00 [4.00–5.00]	87.4%	0.484
En bloc resection rate	5.00 [4.00–5.00]	88.5%	5.00 [5.00–5.00]	90.5%	5.00 [4.00–5.00]	86.8%	0.297
Intraprocedural bleeding rate	4.00 [3.00–5.00]	62.5%	4.00 [3.00–5.00]	65.0%	4.00 [3.00–5.00]	60.3%	0.597
Intraprocedural perforation rate	4.00 [4.00–5.00]	76.0%	4.00 [4.00–5.00]	78.1%	4.00 [3.00–5.00]	73.5%	0.941
Postprocedural bleeding rate	4.00 [4.00–5.00]	77.1%	5.00 [4.00–5.00]	81.0%	4.00 [3.00–5.00]	73.5%	0.196
Postprocedural perforation rate	5.00 [4.00–5.00]	84.7%	5.00 [4.00–5.00]	88.3%	5.00 [4.00–5.00]	82.1%	0.244
Dissection speed	4.00 [3.00–4.00]	53.1%	4.00 [3.00–4.00]	51.1%	4.00 [3.00–4.00]	55.0%	0.860

Supplementary Table 1 provides an overview of training modalities, quality metrics and personal considerations with a gender-based analysis.

Macro-area based differences in ESD training and resources

Global distribution of ESD centers in our sample by macro-area is shown in [Fig. 2], with the highest proportion of HV centers located in Northwest Europe and Asia-Pacific, whereas Africa, the Middle East, and Southeast Europe more frequently hosted LV centers (P < 0.001).

Fig. 2 Map of respondent countries showing prevalence of high-volume ESD centers. Countries not represented in the survey are shown in gray. Color intensity corresponds to increasing proportions of respondents originating from high-volume centers.

Routine tracking of ESD quality indicators also showed marked geographic variation, being most common in Northwest Europe, whereas Africa and the Middle East, Southeast Europe, and America reported substantially lower implementation rates (P = 0.04).

Adherence to the ESGE/ASGE curriculum varied considerably across macro-areas. Among those who had completed a program fulfilling the ESGE/ASGE core curriculum (overall rate of 43.0%), the highest proportions were observed in Northwest Europe (69.1%), followed by Africa and the Middle East (47.6%) and America (34.8%), whereas the lowest were in Southeast Europe (29.4%) and Asia-Pacific (35.7%) (P < 0.001).

Median years of ESD training ranged from 2 (1–5) in Africa and the Middle East to 6 (3–10) in Asia-Pacific and America, with Northwest Europe [3 (2–8)], and Southeast Europe [3 (2–6)] showing intermediate values (P = 0.006).

Similarly, median duration of ESD experience also differed significantly between regions (P < 0.001), ranging from (1 [0–5]) years in Africa and the Middle East to (5 [3–6]) years in America and in Asia-Pacific (5 [2–10]) to intermediate values in Northwest Europe (3 [2–7]), and Southeast Europe (3 [1–6]).

High course or fellowship costs was the most frequently cited barrier in several regions, with marked disparities: 66.7% of respondents in Africa and the Middle East and 47.1% in Southeast Europe identified this as a significant limitation, compared with only 16.2% in Northwest Europe (P < 0.001).

Limited available time for training differed significantly across regions, being most frequently rated as a major barrier (score ≥ 4) in Asia-Pacific (55.3%) and Southeast Europe (48.6%), followed by Africa and the Middle East (41.7%) and America (36.0%), whereas the lowest proportion was observed in Northwest Europe (21.0%) (P < 0.001).

Access to training models also varied (P = 0.082), with no access most frequently reported in Africa and the Middle East (57.1%) and Southeast Europe (54.9%) and least in America (26.1%). Free access was uncommon, ranging from 5.9% in Southeast Europe to 20.6% in Northwest Europe, whereas limited access ranged from 28.6% to 52.2%.

Willingness to relocate for training was generally high, with 41.0% willing to move and another 45.3% willing, depending on funding and location. Africa and the Middle East had the highest proportion fully willing to relocate (52.4%), whereas America showed the highest proportion unwilling (26.1%), although differences were not statistically significant (P = 0.086).

Discussion

This international survey provides the first global snapshot of ESD training, exploring current practice, perceived barriers, and adherence to institutional curricula from the perspectives of both trainees and trainers.

Our study showed marked variation in training design, procedure exposure, and resource availability across settings, with potential consequences for competency development and standardization.

In particular, adherence to the International Societies curriculum was lowest in LV centers, consistent with endoscopic retrograde cholangiopancreatography and other advanced endoscopy data showing that HV centers are more likely to adopt formalized training [8]. A significant trainer-trainee difference (P = 0.001) suggests gradual progress towards greater standardization.

Use of ex vivo and in vivo models was limited, particularly outside the Asia-Pacific region. This contrasts with ESGE recommendations and the Japanese model, where simulator and animal training precede patient-based procedures and are linked to improved safety and higher R0 resection rates [9]. Broader access to effective platforms, including low-cost artificial models, could help bridge the gap between LV and HV centers and promote more uniform competency across regions [10].

Geographic disparities in training infrastructure were also pronounced in our sample. Northwest Europe and Asia-Pacific had the highest proportion of HV centers, patterns driven mainly by countries with strong ESD traditions such as France and Japan, respectively. In contrast, Africa/Middle East and Southeast Europe were predominantly LV. Given that procedure volume is a critical determinant of technical proficiency, these patterns likely contribute to regional variability in core curriculum adherence. Such discrepancies may reflect differences in healthcare organization, training culture, and relative prioritization of ESD compared with other therapeutic modalities [11].

Barriers to training were similarly region-dependent. High course or fellowship costs were most frequently reported in Africa, the Middle East (66.7%), and Southeast Europe (47.1%), consistent with previous studies identifying financial constraints as a significant limitation to advanced endoscopy training [12]. Trainees from LV centers showed less willingness to self-fund their training, suggesting that economic barriers may limit access even when motivation exists. Increased institutional support is needed to ensure more equitable training opportunities.

Access to training models was uneven: These findings echo previous reports that simulator-based ESD training remains patchy worldwide. Expanding use of low-cost artificial, synthetic simulators and even remote training could substantially improve access, reduce dependence on live animal models, and facilitate early-stage skill acquisition, especially in resource-limited settings [13] [14].

Regarding technical training, clear priorities emerged without substantial differences between trainees and trainers. ESU training appears particularly relevant, likely due to heterogeneity of settings across referral centers [15]. Intensive courses, often held in HV centers, provide concentrated exposure and expert supervision, explaining their high perceived value.

Higher ratings from trainees, especially those from MV and HV centers, for traction and underwater techniques may reflect greater exposure to innovations during training in expert centers and a greater willingness to adopt new procedure strategies. Trainers, with established expertise in conventional methods, may place less emphasis on newer modalities that are more commonly encountered in HV settings [16] [17].

Quality indicator monitoring was reported in only 62.5% of centers. This aligns with previous reports showing variable adoption of audit and benchmarking systems in therapeutic endoscopy [18]. Notably, dissection speed was rated lowest, supporting the principle that completeness and safety outweigh procedure time in competency assessment.

On the contrary, there was no clear consensus on the number of procedures required to define trainee independence: 42.7% of respondents indicated 10 to 25 ESDs, whereas 32.6% suggested 26 to 50. This variability underscores the need for broader implementation of the ESGE curriculum, which recommends a minimum of 10 supervised human cases before independent practice and at least 25 ESDs per year to maintain procedure competence.

A marked gender imbalance emerged, extending beyond overall participation to professional roles, with women underrepresented among trainers and reporting lower cumulative endoscopic and ESD experience. Consistent with previous reports in interventional endoscopy [19] [20] this pattern suggests presence of structural barriers along the career pipeline rather than differences in professional interest or aptitude. Notably, gender-stratified analyses (Supplementary Table 1) showed no meaningful differences in perceived training barriers, preferred training modalities, or prioritization of quality indicators, indicating broadly comparable training expectations across genders.

Although greater gender diversity has been associated with improved healthcare outcomes [20], our findings imply that reducing gender disparities in ESD training will require addressing organizational and structural determinants—such as equitable access to advanced training opportunities, transparent fellowship selection processes, and sustained mentorship.

Our study has several strengths. To our knowledge, this is the first survey offering a worldwide overview of ESD training, with responses from 39 countries that allow meaningful geographic comparisons and highlight region-specific needs.

In addition, it captures perspectives of both trainees and trainers, allowing for an analysis of perception differences based on training status.

Finally, the questionnaire explored a wide range of domains, including training structure, technical strategies, quality metrics, and perceived barriers, thereby offering a comprehensive assessment of current practice and adherence to training programs.

However, some limitations should be acknowledged. Use of self-reported data introduces potential for recall bias and subjective interpretation of survey items. Geographic representation was uneven, with underrepresentation of some regions known to host HV ESD centers, such as East Asia (e.g. China and Korea), as well as macro-areas with relatively small numbers of respondents, including Africa and the Middle East. Consequently, comparisons across regions should be interpreted cautiously. Finally, adherence to ESGE/ASGE ESD training curricula was self-reported and reflects fulfilment of recommended structural and training requirements rather than an objective assessment of procedure competence, because meeting curriculum criteria does not necessarily equate to technical proficiency.

Conclusions

In conclusion, these findings highlight areas where efforts toward standardizing ESD training may be needed across the surveyed regions, with prioritization of equitable access to high-quality learning opportunities, particularly through intensive short-term courses that may help bridge geographic and institutional disparities while complementing lengthy and costly fellowships.

Contributorsʼ Statement

Giulio Calabrese: Conceptualization, Data curation, Formal analysis, Methodology, Writing - original draft, Writing - review & editing. Marcello Maida: Conceptualization, Data curation, Methodology, Writing - review & editing. Shaimaa Elkholy: Data curation, Writing - review & editing. Manol Jovani: Data curation, Writing - review & editing. David James Tate: Methodology, Writing - review & editing. Hugo Uchima: Data curation, Writing - review & editing. Yohei Minato: Data curation, Writing - review & editing. Partha Pal: Data curation, Writing - review & editing. Sunil Gupta: Data curation, Writing - review & editing. Rui Morais: Data curation, Writing - review & editing. Torsten Beyna: Data curation, Writing - review & editing. Jeremie Jacques: Conceptualization, Data curation, Methodology, Supervision, Writing - review & editing. Sandro Sferrazza: Conceptualization, Methodology, Supervision, Writing - review & editing.

Conflict of Interest

The authors declare that they have no conflict of interest.

Acknowledgement

We acknowledge the ESD Training Survey Study Group: Taiymi Afafe, Hasan Maulahela, Alberto Murino, Renata Nobre, Darshan Parekh, Yuichi Shimizu, Andrea Sorge, Andrei Voiosu, and Tobias Zuchelli.

• References

• 1 Pimentel-Nunes P, Libânio D, Bastiaansen BAJ. et al. Endoscopic submucosal dissection for superficial gastrointestinal lesions: European Society of Gastrointestinal Endoscopy (ESGE) Guideline - Update 2022. Endoscopy 2022; 54: 591-622
• 2 Yamamoto Y, Fujisaki J, Ishiyama A. et al. Current status of training for endoscopic submucosal dissection for gastric epithelial neoplasm at cancer institute hospital, Japanese Foundation for Cancer Research, a famous Japanese hospital. Digest Endosc 2012; 24: 148-153
• 3 Pimentel-Nunes P, Pioche M, Albéniz E. et al. Curriculum for endoscopic submucosal dissection training in Europe: European Society of Gastrointestinal Endoscopy (ESGE) Position Statement. Endoscopy 2019; 51: 980-992
• 4 Aihara H, Dacha A, Anand GS. et al. Core curriculum for endoscopic submucosal dissection (ESD). Gastrointest Endosc 2021; 93: 1215-1221
• 5 Schlachterman A, Yang D, Goddard A. et al. Perspectives on endoscopic submucosal dissection training in the United States: A survey analysis. Endosc Int Open 2018; 6: E399-E409
• 6 Alfarone L, Schaefer M, Wallenhorst T. et al. Impact of annual case volume on colorectal endoscopic submucosal dissection outcomes in a large prospective cohort study. Am J Gastroenterol 2025; 120: 370-378
• 7 United Nations. Department of Economic and Social Affairs. Statistics. UNSD Publications Catalogue. https://unstats.un.org/UNSDWebsite/Publications/PublicationsCatalogue/109
• 8 Hamesch K, Cahyadi O, Dimitriadis S. et al. Endoscopic retrograde cholangiopancreatography training conditions, results from a pan-European survey: Between vision and reality. United European Gastroenterol J 2025; 13: 474-487
• 9 Oda I, Odagaki T, Suzuki H. et al. Learning curve for endoscopic submucosal dissection of early gastric cancer based on trainee experience. Digest Endosc 2012; 24: 129-132
• 10 Yzet C, Jacques J, Lafeuille P. et al. Does development of submucosal dissection models influence quality of training? Comparison of existing models. Endosc Int Open 2025; 13: a26215244
• 11 Hadjinicolaou AV, Pappas A, Sujendran V. et al. Untutored training pathway to achieve competence in esophagogastric endoscopic submucosal dissection in a Western center. Gastrointest Endosc 2024; 99: 439-443.e6
• 12 Thomson S, Hair C, Oyeleke GK. Outside the training paradigm: challenges and solutions for endoscopy provision in resource-limited settings. Techniq Innovat Gastrointest Endosc 2024; 26: 270-282
• 13 Mitsui T, Sunakawa H, Yoda Y. et al. Novel gastric endoscopic submucosal dissection training model enhances the endoscopic submucosal dissection skills of trainees: A multicenter comparative study. Surg Endosc 2024; 38: 3088-3095
• 14 Ono K, Ohata K, Ide D. et al. Potential for remote hands-on training system for colorectal endoscopic submucosal dissection. VideoGIE 2025; 10: 428-433
• 15 Sferrazza S, Calabrese G, Maselli R. et al. Underwater techniques in gastrointestinal endoscopy: diving into the depths. Cancers (Basel) 2024; 16: 3535
• 16 Van Der Voort V, Legros R, Albouys J. et al. Multi-strategy endoscopic submucosal dissection of a challenging lesion. Endoscopy 2025; 57: E114-E115
• 17 Maida M, Facciorusso A, Marasco G. et al. Comparative efficacy of different techniques for colonic endoscopic submucosal dissection: A network meta-analysis of randomized controlled trials. Digest Liver Dis 2025; 57: 527-534
• 18 Fuccio L, Bhandari P, Maselli R. et al. Ten quality indicators for endoscopic submucosal dissection: what should be monitored and reported to improve quality. Ann Transl Med 2018; 6: 262
• 19 Kotha S, Long M, Berry P. Gender disparity in hepatobiliary endoscopy training and delivery: Results of a nationwide survey. Endosc Int Open 2025; 13: a25056019
• 20 Yu JX, Enestvedt B, Anderson MA. et al. Gender disparities in advanced endoscopy fellowship. Endosc Int Open 2021; 9: E338-E342

Publication History

Received: 10 October 2025

Accepted after revision: 12 February 2026

Accepted Manuscript online:
16 February 2026

Article published online:
06 March 2026