This manuscript represents the outcome of a formal Delphi process resulting in an
official Position Statement of the ESGE and provides a framework to develop and maintain
skills in EMR. This curriculum is set out in terms of the prerequisites for training,
the theoretical knowledge and practical skills required for completion of training,
and how competence should be defined and evidenced prior to independent practice.
Abbreviations
CT:
computed tomography
DMI:
deep mural injury
EMR:
endoscopic mucosal resection
ENTS:
endoscopic non-technical skills
ESGE:
European Society of Gastrointestinal Endoscopy
GI:
gastrointestinal
GPAT:
Global Polypectomy Assessment Tool
GRADE:
Grading of Recommendations Assessment, Development and Evaluation
JNET:
Japan NBI (narrow-band imaging) Expert Team
KPI:
key performance indicator
LNPCP:
large (≥ 10 mm) nonpedunculated colorectal polyp
MDT:
multidisciplinary team
NICE:
Narrow-band imaging International Colorectal Endoscopic classification
SMI:
submucosal invasion
SMIC:
submucosal invasive cancer
SMSA:
size, morphology, site, and access
SSL:
sessile serrated lesion
STSC:
snare-tip soft coagulation
TTS:
through the scope
Definitions
Cold snare polypectomy
Colorectal polypectomy without the use of electrosurgical energy
Competency framework
A set of statements specifying the standards required of a practitioner deemed competent
in the specified technique
Competent EMR practitioner
A fully trained practitioner of EMR as per the competencies defined in this curriculum
Practitioner credentialling
Process of defining that a particular endoscopic practitioner has the knowledge and
technical expertise to perform (specific levels of) colorectal polypectomy
Delayed perforation
A clinical syndrome of pain and raised inflammatory markers after polypectomy suggesting
perforation and confirmed by imaging
Demarcated area
An area within an LNPCP where a regular pit/vascular pattern becomes disordered
Electrosurgical unit
A generator capable of delivering electrosurgical energy
EMR
Colorectal polypectomy of an LNPCP using a submucosal injectate and a snare
Hot snare polypectomy
Colorectal polypectomy using electrosurgical energy
Intraprocedural bleeding
Bleeding during an EMR procedure which requires endoscopic control methods
Intraprocedural perforation
Target sign (Sydney DMI type III) or actual hole (Sydney DMI type IV/V) detected during
a polypectomy procedure
Low risk submucosal invasive cancer (histologic diagnosis)
< 1000-µm depth of invasion measured from the muscularis mucosae, well or moderately
differentiated, no tumor budding, no lymphovascular invasion
Post-resection defect
The area of exposed submucosa remaining after EMR
Snare pivot technique
(Starting with an open snare) using the tip of the snare to push the mucosa away allowing
the open snare to be directed over the polyp tissue even if not positioned at 6 o’clock
Introduction
Endoscopic mucosal resection (EMR) was first referred to by Dehyle in 1973 as a technique
to remove rectal polyps transanally [1 ] and later, in Japan, as a method for removing early gastric cancer [2 ]. It describes any diathermic resection of a gastrointestinal (GI) lesion using a
snare with prior submucosal injection [3 ]
[4 ] (hot snare EMR). Over the years this definition was broadened to involve polyp resections
that did not require electrosurgical energy (cold snare EMR) [5 ].
EMR is the preferred technique to remove flat and sessile ≥ 10-mm polyps in the colon.
It is safe and effective [4 ]
[6 ], with significantly lower cost [7 ], morbidity, and mortality [8 ] compared with surgical resection. In expert hands, more than 98 % of colorectal
polyps can be removed completely using EMR [4 ]
[6 ]
[9 ]
[10 ], without resorting to more resource-intensive endoscopic [11 ]
[12 ] or surgical techniques.
Despite its widespread applicability, training in EMR is often experiential, unstructured,
and dependent on trainers without conscious competence in the technique [13 ]. Compounding this, there is no curriculum for how to train in the procedure and
differing published best practice advice on how to perform high quality procedures.
Aims
With this curriculum, the European Society of Gastrointestinal Endoscopy (ESGE) aims
to promote a standardized evidence-based approach to the practice of EMR, including
training. Of course, there are many ways to approach any clinical problem and neither
curricula nor guidelines should ever be a substitute for clinical judgement.
The curriculum is split into pre-adoption and endoscopy unit requirements for the
performance of EMR; minimum required theoretical knowledge; techniques and approach
before, during, and after EMR; and potential key performance indicators (KPIs) for
lifelong learning. A competency framework for training (the Global Polypectomy Assessment
Tool [GPAT]) is proposed at the end of this document in an attempt to standardize
both the language used in polypectomy training and the assessment of competency.
Methods
This curriculum was developed through a Delphi consensus process amongst international
experts in EMR [14 ]. As chair of the ESGE curricula working group, R.B. invited D.J.T. to be the section
chair for the EMR training curriculum. After a call for participants in January 2020,
R.B. and D.J.T. selected the members, based on the planned curriculum, and their EMR
experience and publications, in March 2020.
The term “endoscopic mucosal resection” in this curriculum relates exclusively to
the endoscopic resection of large (≥ 10 mm) or laterally spreading nonpedunculated
colorectal polyps (LNPCPs) using a snare with prior submucosal injection, with or
without electrosurgical energy.
Because the topic of the curriculum pertains to an area with very little published
evidence, an initial Delphi consensus was first undertaken to define a structure for
the curriculum based on three questions [15 ].
A What are the pre-adoption requirements to start EMR training?
B What are the training/learning steps to achieve competence in EMR?
C What are the assessment criteria for EMR proficiency (being competent and maintaining
competence)?
A total of 11 international expert practitioners of EMR were invited to participate,
based upon their international reputation (for training in EMR and hosting internationally
recognized training courses in EMR), publications, and personal motivation (see Acknowledgments).
The experts were asked to provide open-ended responses to 35 questions covering the
following domains regarding training in EMR (pre-adoption requirements, endoscopy
unit requirements, theoretical knowledge, pre-procedure requirements, competencies
during EMR, post-EMR requirements, requirements for completion of training, and lifelong
KPIs) (the original questions are detailed in Appendix 1 s, see online-only Supplementary material). Experts formulated responses to these initial
questions using the PICO approach (where P stands for population/patient/problem,
I for intervention/indicator, C for comparator/control, and O for outcome).
A first round of anonymous online voting on the categorized statements was performed
by the experts before the first meeting of the taskforce in May 2020. All rounds of
voting were based on a 7-point Likert scale ranging from “Strongly disagree” (1) to
“Strongly agree” (7). During the first taskforce meeting, all members were given the
chance to discuss and adapt the responses. The group was split into four sub-taskforces
each required to cover two of the above domains.
During the adaptation phase, taskforce members conducted systematic literature searches
based on the PubMed database up until May 2020 using search terms derived from the
statements. Where evidence was found, it was used to adapt the statements and the
strength of evidence rated using the Grading of Recommendations Assessment, Development
and Evaluation (GRADE) system for grading evidence levels and recommendation strengths
[16 ]. Where no evidence was found, the statements were maintained in the form drafted
by experts. Written adaptations were made by the sub-taskforce leader after the first
meeting and submitted to D.J.T., who collated them. The evolution and adaptation of
the clinical statements during the Delphi process was documented.
The second round of voting followed the dissemination of adapted statements. Afterwards,
at a second virtual meeting, statements that did not achieve 80 % consensus were discarded
and further modifications were made to those remaining. At this stage, the GPAT and
the criteria for assessing proficiency at EMR were added and discussed. After the
second meeting, formatting of the statements and further referencing was performed
by the sub-taskforces. A third round of voting decided on those statements that remain
in this final document (at least 80 % agreement).
Important : All the statements in this guideline should be interpreted in the context of the
expectations of a fully competent EMR practitioner after completion of a period of
dedicated training.
1 Preadoption requirements
1 Preadoption requirements
EMR of large nonpedunculated colorectal polyps (LNPCPs) is a complex skill and requires
dedication, a significant time investment, and case experience to reach competency.
1 Competencies prior to EMR training
ESGE recommends that the competencies listed below are accomplished prior to in vivo
EMR training under direct supervision.
Best practice recommendation, no evidence available.
(i ) Competency in diagnostic colonoscopy (as defined by local protocols).
Level of agreement 97 %.
(ii ) Commencement of acquiring theoretical knowledge, such as lesion assessment and decision-making.
Level of agreement 96 %.
(iii ) Competency in < 10-mm polypectomy (with and without electrosurgical energy), pedunculated
polypectomy, and common methods of GI endoscopic hemostasis [17 ] (as defined by local protocols).
Level of agreement 97 %.
2 Standards for endoscopy units offering EMR training
ESGE recommends the minimum desirable standards detailed in the following list be
met by endoscopy units offering training in EMR.
Best practice recommendation, no evidence available.
(i ) Written and/or electronic pre-procedure pathways should exist to facilitate the
assessment of patients with co-morbidity prior to EMR.
Level of agreement 90 %.
(ii ) Written local pre-procedure pathways should exist to direct the correct dosing of
anticoagulants for patients prior to EMR based on international consensus guidelines.
Level of agreement 95 %.
(iii ) Formal procedures should exist, either via written information, remote consultation,
or face-to-face discussion, to communicate specific information to the patient prior
to the procedure, particularly to facilitate excellent bowel preparation.
Level of agreement 96 %.
(iv ) Administrative staff booking complex EMR procedures must have guidance to ensure
adequate time and personnel are allocated for the scheduled procedure relative to
its predicted complexity, perhaps using a validated score, such as the size, morphology,
site, and access (SMSA) polyp score [18 ] for guidance. Extra time should be allocated if training is foreseen.
Level of agreement 96 %.
(v ) Units should have a mechanism in place for the training of nursing staff in the
techniques and equipment required for EMR and in how to assist the proceduralist.
Level of agreement 96 %.
(vi ) There should be a dedicated anesthesia-supported list in the unit performing EMR
for complex and/or predicted protracted EMR cases (e. g. very large resections, predicted
difficult location etc.).
Level of agreement 90 %.
(vii ) All necessary equipment should be available prior to starting an EMR procedure.
Equipment checklists should be available and used within the unit to ensure this is
the case.
Level of agreement 96 %.
(viii ) Image and/or video capture equipment for EMR procedures and the ability to record
these in electronic reporting systems should be available.
Level of agreement 95 %.
(ix ) There should be immediate on-site access to support services (including interventional
radiology and surgery) in the event of a significant bleed or other complication (e. g.
perforation), or a pathway should exist for rapid transfer to a specialist center.
Level of agreement 90 %.
(x ) There should be a dedicated recovery room in the unit able to monitor patients at
a high level for an extended period of time and allow admission to hospital if necessary.
Level of agreement 94%.
(xi ) Written emergency pathways should exist for the swift detection and resolution of
adverse events after EMR.
Level of agreement 97%.
(xii ) There should be access to specialist GI pathologists with an interest in GI neoplasia.
Level of agreement 96%.
2 Theoretical knowledge required for a competent EMR practitioner
2 Theoretical knowledge required for a competent EMR practitioner
3 Assessment of LNPCPS for submucosal invasion
ESGE recommends use of the algorithm shown in [Fig. 1 ] and summarized below to risk assess LNPCPs for submucosal invasion (SMI) prior to
EMR.
Level of agreement 89 %.
Strong recommendation, moderate quality of evidence.
Fig. 1 Illustration of the proposed stepwise approach to cancer detection within colorectal
polyps. First look for surrogate markers of cancer and the presence of a demarcated
area. A demarcated area with a disordered vascular pattern implies the lesion is at
very high or overt risk of cancer and should be considered for en bloc endoscopic
resection or surgery (multidisciplinary decision). If there is no demarcated area
or the demarcated area contains no disordered vascular pattern, the four parameters
of covert cancer risk should be determined to identify the covert risk of cancer and
need for en bloc endoscopic resection.
(i ) Obtain an overall impression of the LNPCP by assessing its full extent.
Consider surrogate markers of SMI including tethering of folds, friability, and ulceration
[19 ]
[20 ].
(ii ) Search for a demarcated area.
A demarcated area is a circumscribed area within a colorectal polyp where there is
a definite change from one pit/vascular pattern to another. If detected perform closer
assessment of the pit/vascular pattern within such an area [21 ] (risk of overt SMI). If disordered (NICE III [22 ]
[23 ], JNET 2B/3 [24 ]
[25 ], Kudo V [26 ]) consider the lesion at very high risk of SMI.
(iii ) If there is no demarcated area detectable (or the demarcated area is not disordered
in pit/vascular pattern), perform a morphologic assessment (Paris classification [27 ], location, size, granularity) to assess the risk of covert SMI [28 ].
In general, larger nongranular lesions in the rectum with large nodular components
are at the highest risk of covert SMI.
Comment Practitioners who wish to train in EMR should also train in optical diagnosis of colorectal
lesions [29 ]. Discrimination of the cancer risk of a given LNPCP is critical to guide accurate
decision-making regarding treatment and to avoid adverse patient outcomes. LNPCPs
at low risk of cancer can be treated safely and effectively with EMR [4 ]
[6 ]. LNPCPs at higher risk of cancer may require more resource-intensive endoscopic
resection techniques [12 ]
[30 ] or surgery [31 ].
An estimate of the specific per-polyp risk of SMI can be obtained using referenced
online scoring systems [32 ]. The threshold of acceptable SMI risk, especially covert risk, which itself determines
the appropriate technique for resection, depends upon multidisciplinary discussion,
local expertise, resource availability, and a discussion between the patient and physician
[33 ].
4 Features prompting consideration of en bloc resction or surgery
ESGE recommends the endoscopic features of an LNPCP detailed in the following list
should prompt consideration of en bloc resection or surgery rather than a piecemeal
approach with EMR ([Table 1 ]).
Strong recommendation, moderate quality of evidence.
(i ) Endoscopic imaging features suggesting low risk superficial SMI (en bloc endoscopic
resection) or deep SMI (surgery) [28 ].
Level of agreement 93 %.
(ii ) Any bulky (Is, IIa+Is with nodule > 10 mm) lesion in the rectum or at the rectosigmoid
junction (high risk for covert SMI) without endoscopic imaging features suggesting
overt SMI (en bloc endoscopic resection) [33 ].
Level of agreement 85 %.
(iii ) Any LNPCP with Kudo Vi/JNET 2B pit/vascular pattern where en bloc resection is not
possible with EMR (en bloc endoscopic resection).
Level of agreement 88 %.
(iv ) Any LNPCP with Kudo Vn/JNET 3/NICE III pit/vascular pattern (surgery).
Level of agreement 92 %.
(v ) Any LNPCP with Paris 0-IIa+c or 0-III morphology and ulceration (surgery).
Level of agreement 91 %.
(vi ) Periappendiceal LNPCPs invading deeply into the appendiceal orifice (defined by
the inability to visualize the entire circumferential margin of the lesion) without
prior appendicectomy (full-thickness resection/surgery) [34 ].
Level of agreement 84 %.
(vii ) Inadequate endoscopic access to the target lesion for EMR even by an expert in the
technique [35 ] (laparoscopically assisted EMR or surgery) [36 ].
Level of agreement 84 %.
Comment In general, risk of SMI is the only evidence-based factor that should prompt a decision
for en bloc resection in place of piecemeal EMR. Residual or recurrent adenoma, given
the efficacy of thermal ablation of the post-EMR margin [37 ]
[38 ], should not be the sole reason to prefer an en bloc approach. It is also possible
to perform piecemeal EMR despite non-lifting (whether in a treatment-naïve or previously
attempted/biopsied LNPCP [9 ]), as long as the risk of cancer (Recommendation 3) is regarded as low.
Table 1
Situations in which en bloc endoscopic resection or surgery may be preferred over
a piecemeal endoscopic mucosal resection (EMR).
En bloc resection (endoscopy) preferred
Surgery preferred
Piecemeal EMR standard of care alternative
Endoscopic imaging suggests superficial submucosal invasive cancer (Kudo Vi, JNET
2B) (see Recommendation 3)
Endoscopic imaging suggests risk of deep submucosal invasive cancer (Kudo Vn / JNET
3/ NICE III) (see Recommendation 3)
All other cases
Bulky LNPCP (Is, IIa + Is with nodule > 10 mm) in the rectum (risk of covert SMI)
LNPCPs with Paris 0-IIa + c morphology without ulceration
LNPCPs invading deeply into the appendiceal orifice (also consider FTR)
Inadequate endoscopic access even by an expert
LNPCP, large nonpedunculated colorectal polyp; FTR, full thickness resection; covert
SMI, invasive cancer not visible on the surface of the polyp.
5 Equipment/materials for EMR
ESGE recommends familiarity with the equipment/materials required for EMR given in
the following list.
Weak recommendation, moderate quality evidence.
(i ) Reasons for the use of carbon dioxide as opposed to air [39 ].
Level of agreement 96 %.
(ii ) Different endoscopes, including pediatric colonoscopes and gastroscopes, and when
to consider their use.
Level of agreement 92 %.
(iii ) Electrosurgical unit, including a clear understanding of the benefits of different
cut and coagulation settings [40 ].
Level of agreement 90 %.
(iv ) Different types of transparent distal attachment caps and their uses.
Level of agreement 90 %.
(v ) Waterjet system and endoscope flushing pump.
Level of agreement 96 %.
(vi ) Injection solutions [41 ], colloid vs. crystalloid [42 ]
[43 ], use of adrenaline (at a preferred dilution of 1:100 000 or less) [6 ], and how to prepare them.
Level of agreement 96 %.
(vii ) Different chromic dyes [44 ] and their properties.
Level of agreement 93 %.
(viii ) A variety of different snares and their uses (different lesions and circumstances).
Level of agreement 93 %.
(ix ) Coagulation forceps [45 ] and their uses.
Level of agreement 92 %.
(x ) Different types of endoscopic clips, including different sizes, brands, and rotatability.
Level of agreement 94 %.
(xi ) Endoscopic tattoo, including technique for placement [46 ].
Level of agreement 94 %.
(xii ) Endoscopic retrieval nets.
Level of agreement 93 %.
6 Minimum reporting requirements
ESGE recommends the minimum reporting requirements detailed in the following list
for EMR procedures.
Strong recommendation, moderate quality of evidence.
(i ) Demographics: name of the patient, date of birth, and one other identifier.
Level of agreement 97 %.
(ii ) Demographics: name of the staff involved, name of the trainee if involved.
Level of agreement 95 %.
(iii ) Demographics: indication for the procedure (including original indication if referral
or procedure with intent of lesion resection).
Level of agreement 96 %.
(iv ) Procedure: drugs used including sedation, anesthesia, antibiotics, and/or antispasmodics.
Level of agreement 98 %.
(v ) Procedure: bowel preparation quality [47 ]
[48 ]
[49 ].
Level of agreement 96 %.
(vi ) Procedure: completeness of the colonoscopy [48 ]
[50 ].
Level of agreement 98 %.
(vii ) Procedure: comfort score if procedure with sedation [48 ]
[51 ]
[52 ].
Level of agreement 87 %.
(viii ) Procedure: photodocumentation of the lesion and defect as a minimum [53 ].
Level of agreement 96 %.
(ix ) Procedure: type of endoscope and specific electrosurgical device (snare/knife etc.)
used.
Level of agreement 89 %.
(x ) Lesion specific (for each lesion/EMR performed): lesion location and size.
Level of agreement 98 %.
(xi ) Lesion specific: ease of access and positioning.
Level of agreement 95 %.
(xii ) Lesion specific: an estimate of position using landmarks, scope tracking, or distance
from the anus during withdrawal with a straight scope.
Level of agreement 93 %.
(xiii ) Lesion specific: description including size, Paris [27 ], granularity [45 ], Kudo pit pattern [54 ]/NICE classification [23 ]/JNET [55 ] if magnification is available.
Level of agreement 99 %.
(xiv ) Lesion specific: previous histology if available.
Level of agreement 95 %.
(xv ) Technical points: injectate used and description of lifting.
Level of agreement 80 %.
(xvi ) Technical points: en bloc resection versus piecemeal resection; if a piecemeal resection
is performed, record 2–4 pieces versus ≥ 5 pieces [56 ].
Level of agreement 82 %.
(xvii ) Technical points: duration of the procedure.
Level of agreement 84 %.
(xviii ) Technical points: difficulties encountered including intraprocedural complications
(perforation, bleeding), if any, and how they were managed.
Level of agreement 96 %.
(xix ) Technical points: description of the mucosal defect at the end of the procedure
[57 ] and if any adjunctive treatment was performed (for instance snare-tip soft coagulation
[STSC] to the margins).
Level of agreement 97 %.
(xx ) Technical points: was the mucosal defect closed with endoscopic clips? And for which
indication – bleeding/perforation?
Level of agreement 97 %.
(xxi ) Technical points: was the defect marked with a tattoo to aid future localization?
Level of agreement 97 %.
(xxii ) Technical points: additional lesions present that were not removed and will need
to be located and resected later.
Level of agreement 98 %.
(xxiii ) Technical points: samples retrieved.
Level of agreement 96 %.
(xxiv ) Fasting instructions and timing of recommencement of clear fluids then normal diet.
Level of agreement 96 %.
(xxv ) Follow-up plan: immediate aftercare, post-procedural medications, and timing for
resumption of anticoagulants.
Level of agreement 96 %.
(xxvi ) Follow-up plan: suggested surveillance interval [58 ] (if pathology results indicate benign disease).
Level of agreement 95 %.
(xxvii ) Follow-up plan: planned further procedures.
Level of agreement 89 %.
(xxviii ) Follow-up plan: any further tests/investigations required.
Level of agreement 88 %.
7 Interpretation of pathology reports
EGSE recommends that the features of the pathologic report concerning the specimen
retrieved after EMR that are detailed in the following list be fully interpretable
by a competent EMR practitioner.
Strong recommendation, moderate quality of evidence.
Correct interpretation of histopathology reports is critical to a complete understanding
of EMR. A practitioner trained in EMR must understand the features of a pathology
report described below because these are important to determine the indication for
earlier follow-up or surgery [59 ]
[60 ]
[61 ].
(i ) General histologic category of the lesion: carcinoma, adenoma – tubular or tubulovillous,
villous adenoma, traditional serrated adenoma, sessile serrated lesion (SSL), hyperplastic
polyp.
Level of agreement 97 %.
(ii ) Degree of cytologic dysplasia – no, low grade or high grade (WHO [62 ]/ Vienna classification [63 ]).
Level of agreement 96 %.
(iii ) Horizontal and deep margin free of polyp (completeness of excision [relevant only
for en bloc resection]).
Level of agreement 96 %.
(iv ) Understand that neither “carcinoma in situ” nor “intramucosal cancer” [63 ] stated in pathology reporting are indications for earlier follow-up or surgery,
both being comparable to high grade dysplasia. Both terms should be interpreted as
high grade dysplasia and managed endoscopically.
Level of agreement 93 %.
(v ) Completeness of piecemeal EMR is a judgement made endoscopically at the time of
the index procedure in conjunction with the results of histopathology. It is confirmed
at surveillance procedures via the endoscopic and/or histologic assessment of the
resection scar.
Level of agreement 94 %.
(vi ) Differentiation of submucosal invasive cancer (SMIC): well, moderately, or poorly
differentiated.
Level of agreement 94 %.
For all LNPCPs containing SMIC
(vii ) Depth of invasion from the muscularis mucosae (< 1000 µm vs. ≥ 1000 µm; sessile
polyp) [64 ]
[65 ].
Level of agreement 90 %.
(viii ) Presence of lymphovascular invasion [59 ]
[60 ]
[61 ].
Level of agreement 94 %.
(ix ) Presence of tumor budding [66 ]
[67 ].
Level of agreement 94 %.
Comment For LNPCPs, the depth of submucosal invasion has been classified using the level
system [68 ]
[69 ], which divides the submucosal layer into sm1, sm2, and sm3. The classification cannot
however be applied when lesions have been resected endoscopically, as the full thickness
of the submucosal layer is not usually included. Using a measurement of the distance
of invasion from the muscularis mucosae was proposed as an alternative [64 ]. Endoscopically resected LNPCPs containing SMIC with a depth of invasion from the
muscularis mucosae < 1000 μm without any other histologic risk factors can usually
be followed up without additional surgery, including lymphadenectomy, as the risk
of metastasis is almost nil in such cases [65 ].
8 Attendance at multidisciplinary meetings
ESGE recommends that trainees in EMR regularly attend a multidisciplinary meeting.
Level of agreement 93 %.
Strong recommendation, moderate quality of evidence.
Comment Multidisciplinary team (MDT) meetings have become the standard of care in the field
of oncology. Such meetings usually comprise of a surgeon, oncologist, radiotherapy
specialist, radiologist, and histopathologist. There is a substantial body of evidence
to support the efficacy of such MDT meetings in terms of important outcomes for colorectal
cancer patients (including patient survival) [70 ]
[71 ]
[72 ]
[73 ]. An MDT approach can also facilitate appropriate case selection of early colorectal
cancer for endoscopic treatment, including EMR [74 ].
Trainee inclusion in an MDT is recommended if they are routinely practicing endoscopic
resection in the lower GI tract, both to identify lesions that can be handled endoscopically
and to understand the referral, pre-assessment, and treatment pathways for the complete
range of lesions that may present to them.
3 Before EMR
9 Ground rules prior to starting emr training
ESGE recommends that, prior to starting training and the establishment of a trainee–trainer
relationship, the ground rules listed below should be established.
Best practice recommendation, low quality evidence.
(i ) Endoscopists who provide training in EMR (EMR trainers) should have fulfilled the
requirements in this competency framework (Recommendation 37), be aware of their own
outcomes, and follow them up (Recommendation 38).
Level of agreement 86 %.
(ii ) Trainees in EMR should first undertake a period of observation: watching experts
perform procedures and verbalizing thoughts on lesions and the approach to their resection.
The number of procedures is not specified and should be agreed during discussion between
the trainee and the trainer.
Level of agreement 92 %.
(iii ) During this observation period, the relevant literature on endoscopic resection
should be thoroughly reviewed (including the content of Recommendations 3–7).
Level of agreement 91 %.
(iv ) Training in EMR to become competent (as defined in this curriculum) is best accomplished
during a dedicated fellowship of > 6 months.
Level of agreement 88 %.
(v ) During any fellowship, trainees should not initially expect to fully complete all
cases (establishing ground rules is critical). Recommendation 28 may form a structure
for discussing the ground rules of training.
Level of agreement 96 %.
(vi ) The SMSA and SMSA+ scores may be helpful to guide which lesions are suitable for
initial training ([Table 2 ]) [36 ].
Level of agreement 92 %.
(vii ) Training should be competency based, rather than based on specific numbers of procedures,
perhaps focusing on a particular skill per lesion; for example, lesion assessment,
injection technique, snare selection, and placement technique etc. in the manner outlined
in the following sections.
Level of agreement 96 %.
(viii ) Training should commence on lesions of <30 mm in standard colonic locations (not
those described in the SMSA+ score) and progress to larger lesions, difficult locations/difficult
access etc. as skills develop.
Level of agreement 99 %.
(ix ) EMR trainers should ideally have attended a course on the theory of training and
how it applies to training in endoscopy (train-the-trainers course).
Level of agreement 86 %.
Table 2a
Scoring system for the SMSA (size, morphology, site, and access) score[1 ].
Size, cm
Points
Morphology
Points
Site
Points
Access
Points
< 1 cm
1
Pedunculated
1
Left colon
1
Easy
1
1–1.9
3
Sessile
2
Right colon
2
Difficult
3
2–2.9
5
Flat
3
3–3.9
7
> 4
9
SMSA score
1
2
3
4
Total points
4–5
6–9
9–12
> 12
1 The SMSA score [36 ] is composed of four domains, which are allocated points; adding the points for each
domain results in a score of 1–4.
Table 2b
Scoring system for the SMSA+ score[1 ].
Size, cm
Points
Difficult location [2 ]
Points
Non-lifting/ previous attempt
Points
Granularity
Points
< 4
0
No
0
No
0
Granular
0
≥ 4
1
Yes
1
Yes
1
Non-granular ≥ 20 mm
1
SMSA+ score
Positive
Negative
Total points
≥ 1
< 1
1 The SMSA+ score (combines references [75 ]
[76 ]) is binary, either positive or negative. The score is positive if any of the components
are present (and therefore if a score of ≥ 1 is obtained).
2 Difficult location includes direct involvement of the ileocecal valve, or involvement
of a diverticulum, the anorectal junction, the appendiceal orifice, or a flexure.
10 Process for patient consent
ESGE recommends that the aspects of patient consent detailed in the following list
be discussed by a competent EMR practitioner prior to the procedure.
This information should be given to the patient in written format prior to the procedure,
with an opportunity for discussion before commencing bowel preparation.
Strong recommendation, moderate quality of evidence.
(i ) Adequate time should be planned to fully discuss the following risks and benefits
with the patient and allow them to ask questions prior to EMR.
Level of agreement 93 %.
(ii ) Patients should be made aware that, while a significant number of LNPCPs will undergo
malignant transformation [77 ]
[78 ], the timescale of this cannot be predicted. This is particularly relevant if the
patient has a shortened life-expectancy owing to advanced age or co-morbidity.
Level of agreement 94 %.
(iii ) Patients should be made aware of the main advantages and drawbacks of the proposed
endoscopic procedure and its duration. This should be highlighted in the context particularly
of multiple lesions (which may require multiple procedures) and predicted early malignant
lesions [79 ]
[80 ]
[81 ].
Level of agreement 93 %.
(iv ) Patients should be made aware that the alternatives to EMR are other endoscopic
therapies (e. g. endoscopic submucosal dissection), surgery, and endoscopic monitoring,
and should be made aware of the risks/benefits of these.
Level of agreement 99 %.
(v ) Patients should be aware that, while it is commonly possible to perform EMR as a
day-case procedure, there is a small chance that admission may be required for recovery
from sedation (if living alone for example) or an adverse event [6 ].
Level of agreement 94 %.
(vi ) Patients should be aware that there is a small (< 1 %) risk of unrecognized or unresolvable
perforation during EMR and that this may necessitate emergency surgery if it cannot
be managed endoscopically [79 ].
Level of agreement 96 %.
(vii ) Patients should understand that there are some risks associated with the sedation/anesthesia
for their endoscopic procedure. Such risks are generally small for patients who are
well but may be higher for those with substantial co-morbidity.
Level of agreement 97 %.
(viii ) Patients should understand that there is a risk of incomplete resection of their
lesion using EMR. They may therefore need more than one procedure, or even surgery,
to completely remove their lesion [6 ].
Level of agreement 96 %.
(ix ) Patients should be informed that there is an approximately 3 %–12 % risk (primarily
dependent on lesion size, location, and the use of anticoagulants) of significant
bleeding requiring hospitalization or reintervention after EMR, and that this risk
can be reduced, but not fully mitigated, by adjuvant endoscopic techniques such as
clipping [6 ]
[82 ]
[83 ].
Level of agreement 93 %.
(x ) The risk of bleeding associated with anticoagulants should be personalized and a
plan to manage an individual’s anticoagulation communicated in advance of the procedure,
in discussion with other healthcare professionals involved in the individual’s care
and in concordance with published guidance [84 ].
Level of agreement 95 %.
(xi ) Patients should be informed that there is an approximately 0.7 % risk of delayed
perforation requiring surgical intervention after EMR at expert centers [6 ].
Level of agreement 92 %.
(xii ) Patients should be aware that, despite best endoscopic prediction, 5 %–7 % of LNPCPs
that are predicted to be benign from endoscopic pit pattern assessment will contain
occult SMI and further treatment may be required after EMR [4 ]
[28 ].
Level of agreement 92 %.
(xiii ) Patients should be aware that LNPCPs that have undergone previous resection attempts,
including those marked with carbon particle suspension beneath the lesion or previously
biopsied, have a lower rate of successful resection [9 ]
[85 ] and a higher risk of adverse events [9 ] than those that have not undergone previous attempts.
Level of agreement 90 %.
(xiv ) Personal and center-based metrics for EMR (see Recommendation 38) should be available
to patients if available and if requested.
Level of agreement 81 %.
(xv ) Postoperative instructions should be clearly communicated to patients including
what to expect, when to call for help, and a point of contact if problems occur.
Level of agreement 96 %.
(xvi ) Patients should be aware that there is a risk of polyp recurrence after endoscopic
resection, which may require further treatment [6 ]
[86 ]
[87 ].
Level of agreement 95 %.
(xvii ) Patients should therefore be aware of the need for long-term scheduled surveillance
procedures after EMR to detect and treat recurrent polyps and or metachronous lesions
[88 ]
[89 ].
Level of agreement 95 %.
Comment Informed two-stage consent is a crucial part of patient care. The historical approach
of informed consent focusing solely on safety is no longer sufficient. Consent should
include all aspects of the procedure, including other treatment modalities, post-procedure
recovery, future procedures/surveillance dependent on histopathology result etc. Furthermore,
patients should be aware of the precise risk estimation of complications and other
medical outcomes, based on the current literature and the individual center’s experience.
Consent should be obtained by competent EMR practitioners or those in training who
have had their consent process observed.
11 Patient/lesion work-up prior to EMR
ESGE recommends the patient/lesion work-up detailed in the following list should be
conducted by a competent EMR practitioner prior to the procedure.
Weak recommendation, moderate quality of evidence.
(i ) Informed patient consent.
Level of agreement 97 %.
(ii ) The appropriateness of the procedure and the risk of adverse events should be considered
in the context of patient co-morbidities and life expectancy.
Level of agreement 96 %.
(iii ) Relevant specialty review of co-morbid conditions is only needed in specific cases;
most patients do not require a specialist consultation prior to EMR.
Level of agreement 92 %.
(iv ) Anticoagulant and antiplatelet use should be assessed and a plan for adjustment
should be discussed with the patient prior to EMR [84 ]
[90 ].
Level of agreement 96 %.
(v ) The need for prophylactic antibiotics in relation to underlying health conditions
should be determined as per published guidance [91 ]
[92 ].
Level of agreement 89 %.
(vi ) Local guidelines should be followed in preparing patients with an implantable cardioverter–defibrillator,
spinal stimulator, or deep-brain stimulator to undergo EMR.
Level of agreement 93 %.
(vii ) (If the procedure is to be performed under monitored anesthesia care) anesthetist/perioperative
physician review should be arranged as per local guidance.
Level of agreement 85 %.
(viii ) Social factors that may warrant an overnight stay should be assessed prior to booking
the procedure. Alternatively, frail patients living on their own can be asked to stay
with family members, given that overnight hospital stays may not always be practical
owing to local hospital bed constraints.
Level of agreement 92 %.
(ix ) The quality of previous bowel preparation should be discussed when this was previously
poor and the importance of good bowel preparation should be emphasized [93 ]
[94 ].
Level of agreement 95 %.
(x ) Prior radiographic imaging findings relevant to the lesion for resection should
be reviewed by the proceduralist when available; however, radiographic imaging is
generally not required prior to EMR.
Level of agreement 92 %.
(xi ) Patients should be advised not to plan any travel (without emergent access to hospital
emergency services) in the 14 days after EMR.
Level of agreement 82 %.
12 Determining whether a trainee should start an emr procedure
ESGE recommends that the criteria given in the following list should be considered
when determining when/if EMR procedures should be started by a trainee.
Best practice recommendation, low quality evidence.
(i ) From the outset, certain lesions are unsuitable for hands-on training and this should
be recognized by the supervising practitioner (e. g. complex location, very large
lesions, previously attempted lesions, unstable or very co-morbid patient).
Level of agreement 97 %.
(ii ) Prior knowledge of the SMSA and SMSA+ scores ([Table 2 ]) of the target lesion for EMR can be useful to predict the possibilities for training
[36 ]
[75 ]
[76 ].
Level of agreement 93 %.
(iii ) A trainee in EMR should not be paired with a nursing or other colleague training
in assisting at EMR.
Level of agreement 88 %.
13 Determining the resectability of LNPCPS
ESGE recommends that the points given in the following list regarding the resectability
of LNPCPs are considered by a competent practitioner in EMR.
Best practice recommendation, low quality of evidence.
(i ) The endoscopist should make a full and comprehensive assessment of the LNPCP to
be resected including high definition white-light endoscopy, virtual chromoendoscopy,
and use of magnification if available (see Recommendations 3 and 4).
Level of agreement 95 %.
(ii ) LNPCPs suspected of containing deep SMIC are not suitable for attempted EMR.
Level of agreement 94 %.
(iii ) Only SSLs without dysplasia [5 ]
[95 ]
[96 ] are suitable for piecemeal cold snare polypectomy (within the scope of this competency
framework).
Level of agreement 92 %.
(iv ) Photodocumentation of the lesion is strongly recommended prior to starting the endoscopic
resection procedure.
Level of agreement 96 %.
14 Considerations before emr of an LNPCP
ESGE recommends that the points given in the following list should be considered by
a competent EMR practitioner prior to EMR once an LNPCP has been judged endoscopically
resectable.
Best practice recommendation, low quality evidence.
(i ) Adequate colon cleanliness according to the Boston Bowel Preparation scale (BBPS)
[97 ] or another similar scale. BBPS < 2 in the colonic segment containing the LNPCP for
resection suggests inadequate preparation for EMR and an increased risk of failure
and complications [58 ].
Level of agreement 86 %.
(ii ) LNPCPs with endoscopic imaging suggestive of overt superficial SMI or those with
high risk of covert SMI [28 ] should be considered for en bloc resection (see Recommendation 4).
Level of agreement 95 %.
(iii ) An alternative option for large nodules or small areas within LNPCPs suspicious
for SMI is piecemeal resection with en bloc resection of the suspicious area for better
pathology reporting.
Level of agreement 84 %.
(iv ) En bloc EMR for LNPCPs ≥ 15 mm (right colon) and ≥ 20 mm (left colon) should not
always be pursued at all costs when there is no overt and low covert risk of SMI,
owing to the increased risk of intraprocedural complications (e. g. intraprocedural
perforation [98 ]) without benefit in terms of adenoma recurrence [37 ]
[38 ].
Level of agreement 89 %.
(v ) The endoscopist should consider their own capacity to complete the procedure on
that occasion. Credentialling (see Recommendation 37) and the SMSA score ([Table 2 ]) can help in this determination. If there is doubt as to the competency of the endoscopist,
or that the lesion can be fully resected in a single session, and the potential adverse
events of such a resection managed, the endoscopist should not attempt resection and
should consider referring the patient to an EMR specialist.
Level of agreement 96 %.
(vi ) The endoscopist must be able to obtain optimal endoscopic access to the LNPCP under
consideration for EMR (short scope position without looping, allowing one-to-one movement).
Level of agreement 94 %.
(vii ) A plan should be devised for a systematic approach to resection of the LNPCP under
consideration for EMR.
Level of agreement 94 %.
15 Endoscopic non-technical skills
ESGE recommends that a practitioner competent in EMR has developed strong endoscopic
non-technical skills as detailed in the following list.
Best practice recommendation, low quality evidence.
Endoscopic non-technical skills (ENTS) [99 ] concern non-technical, non-theoretical aspects of EMR that are nevertheless critical
for patient outcomes and the well-being of all staff involved in endoscopy procedures.
They can be graded from 1 (poor) to 4 (good) and are described for a competent practitioner
of EMR.
Communication and teamwork
(i ) Exchanging information: gives and receives knowledge and information in a clear
and timely fashion.
Level of agreement 89 %.
(ii ) Maintaining a shared understanding: ensures that both the team and the endoscopist
are working together from the same information and understand the “big picture” of
the case.
Level of agreement 89 %.
(iii ) Maintaining a patient-centered approach: ensures that the patient is at the center
of the procedure, emphasizing safety and comfort, and giving information in a clear
and understandable fashion.
Level of agreement 89 %.
Situational awareness
(iv ) Preparation: ensures that the patient is fit, the procedure is appropriate, and
that it is being performed by an endoscopist with the necessary skills, equipment,
and assistants for safe and successful completion (adequate work-up performed as per
prior Recommendations).
Level of agreement 92 %.
(v ) Continuous assessment: maintains a continuous evaluation of the patient’s condition
and updates the shared understanding to identify any mismatch between the current
situation and expected state.
Level of agreement 92 %.
(vi ) Problem recognition: recognizes a mismatch between the current situation and the
expected state and anticipates what may happen as a result of possible actions, interventions,
or non-intervention.
Level of agreement 92 %.
(vii ) Focus: ensures a lack of distractions and maintains concentration, particularly
during difficult situations.
Level of agreement 92 %.
Leadership
(viii ) Supporting others: provides emotional and cognitive support to team members and
trainees by tailoring leadership and teaching style appropriately.
Level of agreement 91 %.
(ix ) Maintaining standards: supports safety and quality by adhering to current protocols
and codes of clinical practice.
Level of agreement 93 %.
(x ) Dealing with problems: adopts a calm and controlled demeanor when under pressure,
utilizing all resources to maintain control of the situation, and taking responsibility
for patient outcome.
Level of agreement 92 %.
Judgement and decision-making
(xi ) Considering others: generates possible courses of action to solve an issue or problem,
including assessment of risk and benefit.
Level of agreement 89 %.
(xii ) Making decisions: chooses a solution to a problem, communicating this to team members
and implementing it.
Level of agreement 89 %.
(xiii ) Reviewing the situation: reviews the outcomes of the procedure or options for dealing
with problems, reflecting on issues and instituting changes to improve practice.
Level of agreement 88 %.
4 Recommendations during EMR
4 Recommendations during EMR
This section should be read alongside Section 7 Completion of training, which describes
an assessment tool for technique during the performance of EMR. An electronic version
of this curriculum is available at: www.gieqs.com/emr-curriculum-esge . This version contains multiple video demonstrations of the techniques described.
Sections 4.1 and 4.2 are similar, describing the technical approaches to cold and
hot snare polypectomy. Differences are highlighted where present in [Table 3 ]. [Fig. 2 ] depicts best practice and poor practice from a number of the recommendations below.
Table 3
Differences highlighted in this curriculum between the best practice technique for
cold versus hot snare endoscopic mucosal resection (EMR)[1 ].
Domain
Cold EMR statement
Hot EMR statement
Rationale
Injection technique
Approach to lesion lifting
Use injection to improve lesion access and visualization; Recommendation 16(i)
Do not attempt to lift the whole lesion at once. Start by lifting only the part you
will start to resect. Then follow a process of sequential inject and resect for larger
lesions; Recommendation 19(i)
Injection is not required for cold snare EMR; if it is used it is to aid polyp access
or to better define margins.
Snare placement technique
Snare type
Use a dedicated cold snare; Recommendation 17(i)
Use an appropriate snare for the lesion and situation; Recommendation 20(i)
For cold snare EMR, crisper resection margins which are easier to examine for residual
tissue can be achieved using a dedicated cold snare. This is especially important
for piecemeal cold snare polypectomy.
Snare size
A maximum snare diameter of 10 mm is recommended for piecemeal cold snare polypectomy;
Recommendation 17(ii)
Use of an appropriate size snare for the lesion (maximum 15 mm snare for the right
colon and 20 mm for the left colon); Recommendation 20(ii)
Snare sizes > 10 mm do not tend to reliably transect tissue without electrosurgical
energy.116 ]
Maximizing tissue capture
For cold snare using a dedicated thin-wire snare pushing the snare against the colonic
wall is also effective to maximize tissue capture; Recommendation 17(viii)
Use aspiration of luminal gas and firm downward pressure with the tip of the endoscope
whilst closing the snare to aid tissue capture; Recommendation 20(viii)
Use of a dedicated thin-wire snare for cold snare EMR often allows more reliable capture
of tissue without requiring the maneuvers described in Recommendation 20(viii) which
are useful when using a thicker wire snare for hot snare EMR
Snare closure
The endoscopist issues an instruction to cut the tissue and the assistant closes the
snare fully or the endoscopist takes the snare themselves and cuts the tissue. The
assistant should not re-open the snare unless instructed to do so by the endoscopist;
Recommendation 17(xiii)
At this stage the endoscopist may take the snare and close the handle such that approximately
1 cm remains between the fingers holding the snare handle and the thumb. Tactile feedback
helps the endoscopist to detect entrapped muscularis propria and therefore minimizes
the risk of perforation; Recommendation 20(xiii).
Transection during cold snare EMR is usually with the assistant. Unlike hot snare
EMR, the assistant fully closes their hand until the tissue transects because the
technique is dependent upon mechanical pressure of the snare wire on the captured
tissue.
Failed transection
If the tissue does not transect after 5 seconds, the endoscopist should pull gently
on the snare catheter such that the tissue abuts the tip of the endoscope (guillotine
technique) [107 ]; Recommendation 17(xiv).
Check the mobility of captured tissue with the snare closed ensuring it moves independently
to the muscularis propria of the colon; Recommendation 20(xv)
Failed transection in cold snare EMR often describes the situation where too much
submucosa has been captured. In this case gentle traction on the endoscope tip can
solve the problem (guillotine technique). It is possible to capture the muscularis
propria during cold snare EMR so excessive force or adding electrosurgical energy
should be strongly discouraged. In this situation the snare should be re-opened and
the positioning revised.
Inspection of the post-EMR defect
Detection of residual adenoma
Topical application of a chromic dye (or the injectate if it contains a chromic dye)
can be helpful for assessing the defect and the margin after cold snare polypectomy
for residual adenoma [108 ]; Recommendation 18(ii)
Topical application of the injectate (containing a chromic dye) can be helpful for
areas of the defect where there is uncertainty as to the presence of muscularis propria
injury ([Fig. 2 ], images D7 and D8). If the defect stains blue this area represents submucosa and
the muscularis propria is uninjured [108 ]
[120 ]; Recommendation 22(x)
Topical application of a chromic dye can be extremely useful to check for residual
at the margin of a cold/hot snare defect. Muscularis propria injury is extremely rare
in cold snare EMR but in hot snare it should be routinely sought; a chromic dye makes
the delineation between submucosa and muscularis clear
Treatment of bleeding
Bleeding from within a cold snare polypectomy defect does not need treatment unless
pulsatile; Recommendation 18(iv)
The submucosa of the defect should be further inspected for bleeding vessels. These
should be coagulated [118 ]; Recommendation 22(v)
Persistent active bleeding in hot snare EMR suggests a large, transected blood vessel
which should be cauterized (techniques described in Recommendation 29). Immediate
oozing bleeding is common after cold snare EMR but always stops spontaneously if non-pulsatile
Perforation
Be aware that perforation of the muscularis propria after cold snare polypectomy is
extremely rare and described only in case reports [109 ]
[110 ]. This should be treated appropriately if it occurs; Recommendation 18(v)
The muscularis propria should be inspected if visible and described using the Sydney
Classification of Deep Mural Injury [57 ] ([Fig. 2 ], images D7–D9). Any areas suspicious or uncertain for injury to the muscularis propria
(DMI types II–V) should be closed with an endoscopic closure method; Recommendation
22(ix)
Injury to the muscularis propria during cold snare EMR is rare. If muscularis propria
is caught in a snare closed without electrosurgical energy, the snare will not transect.
Gentle traction against the tip of the instrument is acceptable as this will not injure
the muscularis propria. Excessive force should be avoided, as should the use of electrosurgical
energy, because this situation could allow transection of the muscularis propria.
EMR, endoscopic mucosal resection; DMI, deep mural injury; MP, muscularis propria.
1 The resection plane in cold snare EMR is through the muscularis mucosae; in hot snare,
it is the mid-submucosa [122 ]
[123 ]. This explains the differences in required technique and potential adverse events
described above.
Fig. 2 Example images illustrating best practice techniques (with B3,4 and C3,4 , poor practice examples) in endoscopic mucosal resection for: A global competencies; B injection technique; C snare placement technique; D defect assessment technique; showing: A1,2 full appreciation/demonstration of the extent of the polyp to be resected; A3 best positioning with respect to the polyp (at 6 o’clock, close to the colonoscope);
A4 selection of the appropriate technique for the polyp to be resected (e. g. correct
decision for en bloc cold snare polypectomy in this example); A5 tip control (controlled stable and purposeful, resulting in uniform application of
snare-tip soft coagulation); B1,2 injection performed in the correct plane, facilitating access to the lesion; B3 transmural non-dynamic injection (poor practice); B4 non-lifting, with repeated and failed submucosal injection resulting in intramucosal
blebs (poor practice); C1 the snare oriented near to 6 o’clock, visualization of the snare V during closure,
with the snare near to the colonoscope; C2 use of the transected tissue edge as a guide (within the defect), and visualization
of the snare V during closure, with the snare close to the colonoscope; C3 the snare positioned far from the colonoscope, with too much snare extended from
the sheath and no margin of normal tissue (poor practice); C4 poor tissue capture and too far from the colonoscope (likely to result in scraping
of polyp tissue and incomplete mucosal layer excision; poor practice); D1 complete margin ablation, following systematic application with entire margin ablation
achieved; D2,4 incomplete mucosal layer excision with evidence of residual polyp tissue on islands
of muscularis mucosae; D3 residual polyp tissue at the resection margin; D5 evidence of deep mural injury (DMI) type I, with residual polyp tissue at the defect
edge, and failed submucosal injection resulting in an intramucosal bleb; D6 intraprocedural bleeding; D7 DMI type I; D8 DMI type III associated with an area of fibrosis; D9 DMI type IV; D10 complete closure of a post-endoscopic mucosal resection (EMR) mucosal defect to prevent
post-EMR bleeding.
4.1 Best practice technique for EMR (cold snare)
Meticulous and systematic cold snare polypectomy technique is essential to achieve
complete removal of colonic adenomas [5 ]
[95 ]
[96 ]
[109 ]. A recent systematic review and meta-analysis found incomplete resection rates of
17.3 % (95 %CI 14.3 %–20.3 %) for ≤ 10-mm polyps removed by cold snare polypectomy
[97 ]. Moreover, a pooled multicohort retrospective analysis of eight North American studies
estimated that 19 % of interval colorectal cancers were due to incomplete resection
of a previously noninvasive lesion [110 ].
16 Submucosal injection technique
ESGE recommends the best practice submucosal injection technique described below for
cold snare EMR.
Weak recommendation, low quality of evidence.
(i ) Use injection to improve lesion access and visualization ([Fig. 2 ], image B2).
Level of agreement 86 %.
(ii ) Ensure air is expelled from the injection catheter prior to first injection (priming).
Level of agreement 88 %.
(iii ) Injection through clearly benign lesions is acceptable if necessary. Caution should
be exercised when attempting injection through LNPCPs suspected of containing SMIC
because of the risk of malignant seeding [111 ].
Level of agreement 88 %.
(iv ) Injection should be performed dynamically. The injection is started (short stab,
tangential to the mucosa) as the tissue is being punctured and then the catheter is
pulled back and lifted up/away in the desired direction [112 ].
Level of agreement 87 %.
(v ) Avoid intramucosal blebs ([Fig. 2 ], image B4). If these occur, puncture them and re-attempt submucosal injection [113 ].
Level of agreement 87 %.
(vi ) Stop the injection if no lifting is noted (likely transmural/intraperitoneal injection).
Level of agreement 88 %.
(vii ) Obtain an appropriate and sustained lifting of the lesion ([Fig. 2 ], image B2).
Level of agreement 86 %.
(viii ) Do not over-lift, especially in narrow segments (e. g. sigmoid colon) or at flexures.
Level of agreement 85 %.
17 Snare placement, capture, and closure
ESGE recommends the best practice snare placement, capture, and closure technique
described below for cold snare EMR.
Best practice recommendation, low quality evidence.
(i ) Use of a dedicated, thin-wire (0.3-mm wire diameter) cold snare is strongly recommended
when performing piecemeal cold snare polypectomy.
Level of agreement 90 %.
(ii ) A maximum snare diameter of 10 mm is recommended for piecemeal cold snare polypectomy
as tissue transection of pieces > 10 mm regularly leads to stalling of snare transection.
Level of agreement 91 %.
(iii ) Ensure the lesion is positioned as close to the 6-o'clock position as possible (or
is transformed to 6 o’clock using the snare pivot technique).
Level of agreement 91 %.
(iv ) Start at the edge of the lesion.
Level of agreement 92 %.
(v ) Aim to capture a rim of 1–3 mm of normal mucosa ([Fig. 2 ], images A4 and C1). This is critical in cold snare polypectomy owing to the absence
of electrosurgical energy.
Level of agreement 94 %.
(vi ) Ideally the long axis of the snare should be placed parallel to the polyp base.
Level of agreement 93 %.
(vii ) The V of the snare ([Fig. 2 ], images C1 and C2) should be extended just beyond the tip of the snare catheter
prior to snare placement.
Level of agreement 92 %.
(viii ) Use aspiration of luminal gas and firm downward pressure with the tip of the endoscope
whilst closing the snare to aid tissue capture. For cold snaring, using a dedicated
thin-wire snare and pushing the snare against the colonic wall is also effective to
maximize tissue capture.
Level of agreement 92 %.
(ix ) Placement of the snare should be under direct visualization and aligned precisely
with the advancing edge of the mucosal defect (if piecemeal) ([Fig. 2 ], image C2).
Level of agreement 93 %.
(x ) The tip of the snare catheter (and the V of the snare just beyond its tip) should
be close to the endoscope during closure and constantly visualized during closure
if possible ([Fig. 2 ], image C2).
Level of agreement 93 %.
(xi ) The assistant is asked to close the snare and to stop closure, issuing a verbal
alert to the endoscopist, when resistance is felt.
Level of agreement 90 %.
(xii ) At this point, the endoscopist determines whether the captured tissue is what was
desired.
Level of agreement 90 %.
(xiii ) The endoscopist issues an instruction to cut the tissue and the assistant closes
the snare fully or the endoscopist takes the snare themselves and cuts the tissue.
The assistant should not reopen the snare unless instructed to do so by the endoscopist.
Level of agreement 91 %.
(xiv ) If the tissue is not transected after 5 seconds, the endoscopist should pull gently
on the snare catheter so that the tissue abuts the tip of the endoscope (guillotine
technique) [101 ].
Level of agreement 87 %.
(xv ) If the tissue still does not cut despite attempted use of the guillotine technique,
the assistant is asked to slowly open the snare, while continuing gentle traction
on the polyp. This usually leads to the mucosal specimen sliding over a submucosal
protrusion.
Level of agreement 90 %.
(xvi ) If the tissue still is not transected, the snare should be reopened and the procedure
repeated.
Level of agreement 86 %.
(xvii ) We recommend against the addition of electrosurgical energy in the situation where
complete closure of the snare during cold snare resection has not led to transection
in spite of the above measures as this risks injury to the muscularis propria.
Level of agreement 94 %.
(xviii ) Work sequentially across the lesion when performing piecemeal resection, aligning
the snare with the edge of the advancing mucosal defect (to avoid islands).
Level of agreement 91 %.
Comment Use of a dedicated snare allows for clean tissue transection and better analysis
of the expanding mucosal defect. Evidence of increased efficacy is however conflicting.
In a prospective randomized controlled trial examining the removal of polyps ≤10 mm
by cold snare polypectomy, use of dedicated cold snares was associated with a significantly
greater complete resection rate than traditional cold snares (91 % vs. 79 %; P = 0.02) [114 ]. Conversely, in a recent international multicenter randomized trial, use of dedicated
cold snares was not associated with improved complete resection rates [115 ].
18 Inspection of the post-EMR defect
ESGE recommends the best practice technique detailed below for inspecting the post-EMR
defect following cold snare EMR.
Best practice recommendation, low quality evidence.
(i ) Spend an appropriate amount of time inspecting the entire mucosal defect, particularly
the margins, for residual polyp tissue ([Fig. 2 ], images D2–D5). If there are small areas of residual tissue within the defect, cold
snare avulsion can be used to remove these.
Level of agreement 94 %.
(ii ) Topical application of a chromic dye (or the injectate if it contains a chromic
dye) can be helpful for assessing the defect and the margin for residual adenoma after
cold snare polypectomy [102 ] ([Fig. 2 ], images D4 and D5).
Level of agreement 84 %.
(iii ) The use of virtual chromoendoscopy and magnification can be helpful to detect residual
polyp tissue at the defect margin.
Level of agreement 86 %.
(iv ) Bleeding from within a cold snare polypectomy defect does not need treatment unless
pulsatile.
Level of agreement 90 %.
(v ) Be aware that perforation of the muscularis propria after cold snare polypectomy
is extremely rare and described only in case reports [105 ]
[106 ]. This should be treated appropriately if it occurs.
Level of agreement 89 %.
(vi ) Photodocumentation of all areas of the post-EMR defect ([Fig. 2 ], images D1–D5) helps to assess the completeness of the resection. It may also be
useful for medicolegal purposes.
Level of agreement 89 %.
4.2 Best practice technique for EMR (hot snare)
The following references pertain to Recommendations 19, 20, and 21 [45 ]
[53 ]
[81 ]
[116 ]
[117 ]
[118 ]
[119 ]
[120 ].
The submucosal injection technique for hot snare EMR is similar to that for cold snare
EMR; however, endoscopists should avoid attempting to raise the entire lesion at the
outset, especially for large lesions. Instead, a sequential inject-and-resect technique
should be adopted, where only a portion of the lesion is raised, with each submucosal
injection immediately followed by transection of that area.
19 Submucosal injection technique
ESGE recommends the best practice submucosal injection technique described below for
hot snare EMR.
Best practice recommendation, low quality evidence.
(i ) Do not attempt to lift the whole lesion at once. Start by lifting only the part
you will start to resect, then follow a process of sequential injection and resection
for larger lesions.
Level of agreement 91 %.
(ii ) Use injection to improve lesion access and visualization ([Fig. 2 ], image B2).
Level of agreement 91 %.
(iii ) Ensure air is expelled from the injection catheter prior to the first injection
(priming).
Level of agreement 91 %.
(iv ) Injection through clearly benign lesions is acceptable if necessary ([Fig. 2 ], image B1). Caution should be exercised when attempting injection through lesions
suspected of containing SMIC because of the risk of malignant seeding [111 ].
Level of agreement 92 %.
(v ) Injection should be performed dynamically. The injection is started (short stab,
tangential to the mucosa) as the tissue is being punctured and then the catheter is
pulled back and lifted up/away in the desired direction [112 ].
Level of agreement 94 %.
(vi ) Avoid intramucosal blebs (intramucosal injection) ([Fig. 2 ], image B4). If these occur, puncture them and re-attempt submucosal injection [113 ].
Level of agreement 89 %.
(vii ) Stop the injection if no lifting is noted (likely transmural/intraperitoneal injection).
Level of agreement 91 %.
(viii ) Obtain an appropriate and sustained lifting of the lesion ([Fig. 2 ], image B2).
Level of agreement 91 %.
(ix ) Do not over-lift, especially in narrow segments (e. g. sigmoid colon) or at flexures.
Level of agreement 90 %.
20 Snare placement, capture, and closure
ESGE recommends the best practice snare placement, capture, and closure technique
described below for hot snare EMR.
Best practice recommendation, low quality of evidence.
As with cold snare polypectomy, a meticulous and systematic approach to snare resection
is essential to ensure complete resection of colorectal lesions by hot snare EMR.
The differences between the hot snare and cold snare techniques include the recommended
use of a thick-wire snare (wire diameter > 0.4 mm) for hot snare EMR. This further
dictates modifications in the technique owing to the poorer tissue capture properties
of these snares.
(i ) Use of an appropriate snare type for the lesion and situation (e. g. standard thick-wire
snare [wire diameter > 0.4 mm] for hot snare polypectomy).
Level of agreement 92 %.
(ii ) Use of an appropriate size snare for the lesion. Avoid large snares for large lesions
[100 ], with suggested maximums of a 15-mm snare for the right colon and a 20-mm snare
for the left colon (distal to splenic flexure).
Level of agreement 86 %.
(iii ) Ensure the lesion is positioned as close to the 6-o'clock position as possible ([Fig. 2 ], image C1) (by rotation of the endoscope or transformation to 6 o’clock using the
snare pivot technique).
Level of agreement 91 %.
(iv ) Start at the edge of the lesion.
Level of agreement 96 %.
(v ) Aim to capture a rim of 2–3 mm of normal mucosa ([Fig. 2 ], image C1).
Level of agreement 96 %.
(vi ) Where possible, the long axis of the snare should be aligned parallel to the polyp
base.
Level of agreement 89 %.
(vii ) The V of the snare ([Fig. 2 ], images C1 and C2) should be extended just beyond the tip of the snare catheter
prior to snare placement.
Level of agreement 89 %.
(viii ) Use aspiration of luminal gas and firm downward pressure with the tip of the endoscope
whilst closing the snare to aid tissue capture.
Level of agreement 94 %.
(ix ) Placement of the snare should be under direct visualization and aligned precisely
with the advancing edge of the mucosal defect (if piecemeal) ([Fig. 2 ], image C2).
Level of agreement 94 %.
(x ) Placement of the snare should be under direct visualization ensuring a rim of 1–3 mm
of normal tissue ([Fig. 2 ], image C1) completely surrounding the target lesion (if en bloc).
Level of agreement 95 %.
(xi ) The tip of the snare catheter (the V of the snare just beyond its tip) should be
near to the endoscope during closure and constantly visualized during closure ([Fig. 2 ], images C1 and C2) if possible.
Level of agreement 92 %.
(xii ) The assistant is asked to close the snare and to stop closure, issuing a verbal
alert to the endoscopist, when resistance is felt.
Level of agreement 91 %.
(xiii ) At this stage, the endoscopist may take the snare and close the handle so that approximately
1 cm remains between the fingers holding the snare handle and the thumb. Tactile feedback
helps the endoscopist to detect entrapped muscularis propria and therefore minimizes
the risk of perforation.
Level of agreement 88 %.
(xiv ) Placing a marker on the handle of the snare at the position where (during closure
outside the patient) the tip of the snare is seen to enter the sheath may be an alternative
to assess the extent of closure, so that the snare may remain with the assistant.
Level of agreement 86 %.
(xv ) Check the mobility of the captured tissue with the snare closed, ensuring it moves
independently to the muscularis propria of the colon.
Level of agreement 96 %.
(xvi ) Release and reclose the snare if there is concern over capture of the muscularis
propria (assistant or endoscopist suggests too much tissue captured).
Level of agreement 96 %.
(xvii ) Work sequentially across the lesion when performing piecemeal resection, aligning
the snare with the edge of the advancing mucosal defect to avoid islands.
Level of agreement 96 %.
21 Safety checks before applying electrosurgical energy
ESGE recommends the safety checks detailed below be made prior to the application
of electrosurgical energy during hot snare EMR.
Moderately strong recommendation, low quality evidence.
(i ) Use a modern microprocessor controlled electrosurgical unit capable of delivering
a range of currents, including fractionated current [121 ].
Level of agreement 94 %.
(ii ) Ensure the correct settings are used on the electrosurgical unit. It is particularly
important to recheck during longer procedures, especially after each instrument change
or change in patient position.
Level of agreement 93 %.
(iii ) For monopolar devices, ensure that the patient has an electrosurgical plate on and
that the generator is ready to deliver thermal energy.
Level of agreement 93 %.
(iv ) Ensure a stable colonoscope position: if necessary an assistant should support the
scope shaft.
Level of agreement 93 %.
(v ) Ensure hemostatic options are available immediately afterwards if required (e. g.
clips and hemostatic forceps).
Level of agreement 94 %.
(vi ) The lumen should subsequently be inflated to visualize the captured tissue and snare
placement.
Level of agreement 91 %.
(vii ) The endoscopist should perform a mobility check of the captured tissue to ensure
that the muscularis propria is not entrapped.
Level of agreement 90 %.
(viii ) (If the endoscopist is handling the snare) the endoscopist uses tactile feedback
to assess the amount of tissue captured and revises if necessary.
Level of agreement 90 %.
(ix ) (If the assistant is handling the snare) ensure the assistant knows the sequence
of commands on application of electrosurgical energy (e. g. close snare in a controlled
manner to the mark).
Level of agreement 83 %.
(x ) Tent the tissue for resection away from the deeper structures. Apply electrosurgical
energy allowing the energy delivered through the snare to transect the tissue rather
than using hand pressure to avoid cold snare resection. The snare handle should not
snap closed.
Level of agreement 96 %.
(xi ) During the application of fractionated current, beware if the tissue is not transected
within three cycles. Stop further electrosurgical energy application and assess the
situation.
Level of agreement 93 %.
(xii ) Stop the application of electrosurgical energy as soon as the tissue is transected.
Level of agreement 94 %.
22 Inspection of the post-EMR defect
ESGE recommends the best practice technique detailed below for inspecting the post-EMR
defect after hot snare EMR.
Strong recommendation, moderate quality evidence.
(i ) Spend an appropriate amount of time inspecting the entire mucosal defect ([Fig. 2 ], images D2–D5), particularly the margins, for residual polyp tissue [37 ]
[38 ].
Level of agreement 95 %.
(ii ) The use of virtual chromoendoscopy and magnification can be helpful to detect residual
polyp tissue at the defect margin.
Level of agreement 90 %.
(iii ) If removal of the snare is difficult, visible polyp tissue should not be coagulated,
instead it should be removed using a different snare ([Fig. 2 ], image D4) or an avulsion technique prior to considering ablation.
Level of agreement 95 %.
(iv ) After any visible polyp has been removed from the defect margin, a thermal ablation
technique (e. g. STSC; ERBE soft coagulation effect 4, 80 W) should be applied to
the entire margin ensuring a rim of 1–3 mm of complete mucosal ablation. The thermal
ablation technique should only be applied to tissue with a prior submucosal injection
to ensure safety ([Fig. 2 ], image A5).
Level of agreement 88 %.
(v ) The submucosa of the defect should be further inspected for bleeding vessels ([Fig. 2 ], image D6). These should be coagulated [104 ] (see Recommendation 29).
Level of agreement 90 %.
(vi ) The submucosa of the defect should be further inspected for large herniating vessels.
The weight of available evidence suggests that, if vessels are not bleeding, they
do not need to be coagulated [122 ].
Level of agreement 88 %.
(vii ) Significant fibrosis in the submucosal plane should be noted. This may represent
a benign cause such as prolapse, may be an intrinsic feature of a nongranular LNPCP,
or may represent evidence of submucosal invasion [9 ].
Level of agreement 94 %.
(viii ) It should be recognized that no feature (other than bleeding vessels) of the submucosa
of the defect has been shown to predict important outcomes after EMR [104 ].
Level of agreement 90 %.
(ix ) The muscularis propria should be inspected if visible and described using the Sydney
classification of deep mural injury (DMI) [57 ] ([Fig. 2 ], images D7–D9).
Level of agreement 90 %.
(x ) Topical application of the injectate (containing a chromic dye) can be helpful for
areas of the defect where there is uncertainty as to the presence of muscularis propria
injury ([Fig. 2 ], images D7 and D8). If the defect stains blue, this area represents submucosa and
the muscularis propria is uninjured [102 ]
[103 ].
Level of agreement 90 %.
(xi ) Any areas suspicious or uncertain for injury to the muscularis propria (DMI types
II–V) ([Fig. 2 ], images D8 and D9) should be closed with an endoscopic closure method ([Fig. 2 ], image D10), as described in the statement regarding intraprocedural perforation
[57 ].
Level of agreement 95 %.
(xii ) Use of a transparent distal colonoscope attachment can be helpful to fully visualize
a mucosal defect over a fold or at a flexure.
Level of agreement 88 %.
(xiii ) Photodocumentation of all areas of the post-EMR defect helps to assess completeness
of resection ([Fig. 2 ], image D1) and facilitates learning if an adverse event occurs later. It may also
be useful for medicolegal purposes.
Level of agreement 92 %.
23 Application of thermal energy to the defect margin
ESGE recommends the best practice technique detailed below when applying thermal energy
to the defect margin after EMR to reduce adenoma recurrence.
Strong recommendation, high quality evidence.
(i ) The technique for STSC of the post-EMR margin is as follows:
extend the tip of the snare 2–3 mm beyond the catheter sheath
use the scope and the snare tip as one device, do not move the catheter in and out
of the channel
apply pulses of soft coagulation using a light-touch technique to the margin of the
defect, ideally over injected tissue
aim for a rim of 2–3 mm of ablated tissue at the margin
ensure that the margin is ablated around the full circumference without gaps.
Confirmation of the technique is visual identification of the white appearance of
the mucosa after application ([Fig. 2 ], image A5).
Level of agreement 96 %.
Comment Evidence from several randomized studies suggests that thermal ablation of the defect
margin after EMR significantly reduces recurrence at first surveillance colonoscopy,
by approximately 4-fold in expert hands [37 ]. Prospective observational studies [38 ] and meta-analyses [123 ]
[124 ] have confirmed this finding.
24 Alternative methods for removal of non-lifting residual adenomatous tissue
ESGE recommends that, other than snare resection (EMR), the following techniques are
alternative methods for removing non-lifting residual adenomatous tissue.
Strong recommendation, moderate quality evidence.
(i ) Cold avulsion (using a standard biopsy forceps), then thermal ablation using soft
coagulation (ERBE effect 4, 80 W) to the avulsion bed using the snare tip. Non-lifting
tissue should be completely removed prior to application of soft coagulation [9 ].
Level of agreement 84%.
Comment There are many other published techniques to manage residual adenomatous tissue.
Such techniques include, but are not restricted to, hot avulsion [125 ], forced argon plasma coagulation [126 ], and more resource-intensive techniques, such as endoscopic submucosal dissection
[127 ]
[128 ]
[129 ] or full-thickness resection [12 ]. These techniques are outside of the scope of this curriculum and are often not
cost-effective in this situation.
4.3 Best practice for retrieval of polyp tissue after EMR
25 Retrieval of polyp tissue
ESGE recommends the best practice technique described below for retrieval of polyp
tissue after EMR.
Weak recommendation, low quality evidence.
General statements
(i ) Use of an endoscopic retrieval net is recommended when performing piecemeal polypectomy.
Level of agreement 92 %.
(ii ) If performing en bloc resection, the snare can be used to retrieve the resected
polyp tissue.
Level of agreement 88 %.
(iii ) (If using a snare) care should be taken to ensure the snare does not cut through
the lesion and that the scope is used to protect the specimen at retrieval through
the anus by pulling the specimen up against the scope tip.
Level of agreement 86 %.
(iv ) Suction of transected tissue through the endoscope channel for retrieval may interfere
with histologic interpretation.
Level of agreement 84 %.
Technique of retrieval using an endoscopic retrieval net
(v ) The retrieval device should be positioned at 6 o’clock (when using a colonoscope).
Level of agreement 89 %.
(vi ) Pieces should be captured using a sequential place, close, compact, and open approach.
Level of agreement 92 %.
(vii ) All pieces should be captured.
Level of agreement 94 %.
(viii ) Pinning the polyp tissue to a corkboard after en bloc resection is useful, but this
is not useful after piecemeal resection (apart from when a large piece suspicious
for SMI is resected).
Level of agreement 88 %.
4.4 Maximizing the chances of EMR success in difficult cases
26 Strategies for difficult to access LNPCPS
ESGE recommends that the strategies detailed below could be employed to facilitate
access to the target LNPCP where this is difficult.
Moderately strong recommendation, low quality of evidence.
Patient factors
(i ) Change patient position to respect the influence of gravity on fluid/blood (away
from the lesion) and to optimize views.
Level of agreement 96 %.
(ii ) Optimize patient sedation.
Level of agreement 94 %.
(iii ) Administration of antispasmodics.
Level of agreement 93 %.
Insertion challenges
(iv ) Excellent basic colonoscopy technique to arrive at the lesion without looping.
Level of agreement 96 %.
(v ) Use of external pressure to prevent loop formation during insertion.
Level of agreement 86 %.
(vi ) Use of different endoscopes with different lengths, stiffness, and tip-bending properties
etc. (e. g. pediatric colonoscope, gastroscope, balloon-enteroscope).
Level of agreement 93 %.
Poor stability of the endoscope tip
(vii ) Ask assistant/nurse to hold the endoscope as needed.
Level of agreement 90 %.
(viii ) Position the lesion at 5–6 o’clock.
Level of agreement 94 %.
(ix ) Use retroflexion where possible and appropriate.
Level of agreement 91 %.
(x ) Use luminal insufflation and desufflation.
Level of agreement 94 %.
Difficult access
(xi ) Use of a transparent distal attachment (e. g. for flexures, ileocecal valve, or
dentate line).
Level of agreement 91 %.
Polyp characteristics: inadequate lifting or poor snare capture
(xii ) Resection technique: commence resection in a way that facilitates access to the
most difficult part of the lesion early in the procedure to optimize the chance of
complete resection.
Level of agreement 85 %.
(xiii ) Resection technique: use submucosal lifting to promote access to the required part
of the lesion.
Level of agreement 94 %.
(xiv ) Resection technique: use snare resection of the surrounding lifting mucosa to create
a step for snare placement on difficult-to-capture tissue.
Level of agreement 92 %.
(xv ) Resection technique: create a step into the submucosal plane to assist snare placement
(hybrid-EMR).
Level of agreement 85 %.
(xvi ) Resection technique: use an avulsion technique such as cold avulsion.
Level of agreement 85 %.
Suboptimal or incorrect technique
(xvii ) Consider asking for a second opinion from a more experienced colleague at the time
or before a second attempt. Consider that you may not be the best person to perform
the second attempt.
Level of agreement 95 %.
(xviii ) Try a second attempt at EMR prior to surgery (two-stage EMR) [10 ].
Level of agreement 85 %.
Several factors can result in a polypectomy being regarded as “difficult.” These can
be divided into the following seven categories.
Patient factors
In addition to compliance, the subsequent quality of the bowel preparation may also
be affected by patient medication or underlying medical conditions that can affect
GI motility. Poor cognition, obesity, or decreased mobility can limit patient movements.
This may also adversely impact on the decision regarding when to turn a patient to
facilitate reaching or optimizing access to an LNPCP. This may result in a slower
and more challenging resection. Bowel motility and spasm can impair insertion, but
also visualization and access.
Insertion challenges
An endoscopist must first be able to reach the LNPCP before resection can be attempted.
Insertion challenges can result from: fixed segments of colon; long mobile colons
with complex loop formation; patients with a hypersensitive colon and poor tolerance
of any distension or colonic stretch stimulation; or previous abdominal/pelvic surgery.
A range of endoscopic instruments may help mitigate some of these problems (e. g.
a slimmer flexible instrument in a fixed or narrowed colon, a stiffer or longer scope
in redundant loopy colons). The underwater insertion technique is another option for
improving successful intubation, helping to reduce distension and patient discomfort,
improving control of loops and improving lubrication.
Poor stability of the endoscope tip
During resection of an LNPCP, tip control is a key requirement to ensure safe effective
resection. The colonic mucosa should be pleated over the colonoscope to enable a short
and stable position. Ideally the patient should be positioned so that gravity does
not adversely affect stability, whilst maintaining an optimal view of the lesion or
resection field. Endoscopists must employ optimal basic colonoscopy technique during
insertion to reach the lesion, to minimize any potential problems with tip control
during the resection. Assistance may be needed in difficult situations.
Difficult access
LNPCPs can occur anywhere in the colon. Difficult access is associated with flexures,
folds, the ileocecal valve, terminal ileum, or being located in the orifices of the
appendix or a diverticulum. Changing the patient’s position (often multiple times)
or using a distal tip attachment may help improve access to the lesion. Some lesions
may be best visualized or resected in part or entirely in a retroverted view.
Inadequate lifting or poor snare capture
The presence of fibrosis secondary to disease or previous intervention can increase
the difficulty of resection. Pre-resection biopsy can also contribute to submucosal
fibrosis [130 ]). Additionally, the presence of submucosal fat can add to the technical challenges
of resection in the submucosal plane. This is particularly noticeable for lesions
in the right colon, around the ileocecal valve.
Polyp characteristics
Specific types of LNPCPs can have an increased association with fibrosis, in particular
nongranular LNPCPs [9 ]. A larger exophytic component of a lesion is associated with an increased size in
feeding vessels. The distal rectum is also a highly vascular area and increased care
to ensure hemostasis during resection is needed for polyps in this location.
Suboptimal or incorrect technique
Resection technique needs to ensure safety, with resection being oriented through
the submucosal plane and parallel to the underlying muscular propria. If there is
inadequate lifting or poor orientation, the risk of damage to the muscular propria
increases. Incorrect electrosurgical energy application can result in a risk of immediate
or delayed perforation and/or bleeding. To ensure complete resection, a normal cuff
of mucosa should be removed with the edge of the lesion and any tissue bridges should
be avoided. Adjuvant post-resection techniques should be employed, where appropriate,
to reduce the potential for complications, including recurrence.
27 Strategies for EMR in specific locations
ESGE recommends the strategies or maneuvers described below be used to assist in EMR
of lesions in specific locations.
Strong recommendation, moderate quality of evidence.
(i ) The position of the patient is critical, using gravity to maximize the luminal view
and access to the lesion. Consider the right lateral position for lesions in the sigmoid,
descending colon, splenic flexure, and cecal pole; the left lateral position for lesions
in the hepatic flexure and ascending colon; the supine position for lesions in the
transverse colon; and the prone position if other positions are difficult in a patient
with a long redundant colon, particularly for lesions in the rectosigmoid [131 ].
Level of agreement 88 %.
(ii ) Ileocecal valve: use a distal attachment. Assess the lesion carefully prior to resection.
Be aware that risk factors for failed endoscopic resection are: deep ingrowth into
the ileum and circumferential involvement of the ileocecal valve lips. Start by using
dynamic injection to encourage the ileal margin of the lesion into the colonic lumen.
Start resection from this margin. A small stiff snare often helps at this location
[132 ].
Level of agreement 92 %.
(iii ) Appendiceal orifice: lesion assessment is critical. Risk factors for failure of
endoscopic resection include invasion of the lesion beyond the endoscopic view into
the appendiceal orifice and > 50 % circumferential involvement of the appendiceal
orifice. A small stiff snare often helps in this location [34 ], as may use of the underwater technique [133 ].
Level of agreement 92 %.
(iv ) Rectosigmoid: consider use of a gastroscope, which will offer the ability for retroflexion.
Level of agreement 92 %.
(v ) Flexures: use of distal attachments may help.
Level of agreement 93 %.
(vi ) Anorectal junction: use a distal attachment. Use a small snare to commence resection
at the anal verge, allowing expansion of the mucosal defect into the rectum and facilitating
further resection. Use local anesthetic as part of the injectate in this area to provide
analgesia after the procedure [92 ]
[134 ].
Level of agreement 91 %.
4.5 Training aspects related to EMR procedures
28 Determing whether EMR procedures should be taken over by the supervising practitioner
ESGE recommends the criteria listed below should be considered when determining when/if
trainee EMR procedures should be taken over by their supervising practitioner.
Best practice recommendation, low quality evidence.
The following statements assume the supervisor has insight into the trainee’s knowledge
and skills. It may be very difficult to apply this approach for a trainee not known
to the trainer.
Global factors
(i ) Trainee is unable to carry out the instructions of the trainer despite comprehending
them.
Level of agreement 95 %.
(ii ) Lack of adequate progress due to any reason relevant to the level of training or
competency.
Level of agreement 94 %.
Patient factors
(iii ) Patient withdraws consent for a training procedure to be performed.
Level of agreement 90 %.
(iv ) Patient is in significant discomfort and this is thought to be due to the involvement
of a trainee.
Level of agreement 93 %.
(v ) Patient becomes medically unstable.
Level of agreement 94 %.
Technical factors
(vi ) Suspicion of area(s) of SMI within the lesion for EMR, which are deemed not to have
been sufficiently/adequately interrogated prior to the EMR or by the trainee, and/or
where a precise en bloc resection is required that may not be achievable by the trainee.
Level of agreement 95 %.
(vii ) Unstable colonoscope position, with the trainee not able to maintain the optimal
position on the lesion for snare placement (5–6 o’clock) despite input from the trainer.
Level of agreement 94 %.
(viii ) Trainee unable to obtain adequate submucosal lift.
Level of agreement 93 %.
(ix ) Trainee repeatedly cold snares the target lesion or resections lead only to superficial
scraping of polyp tissue and incomplete excision of the mucosal layer.
Level of agreement 94 %.
Adverse events
(x ) Severe intraprocedural bleeding if the trainee is not competent/able to address
this.
Level of agreement 94 %.
(xi ) Perforation if the trainee is not competent/able to address this.
Level of agreement 95 %.
(xii ) Trainee requests that the supervisor takes over the procedure.
Level of agreement 92 %.
Time pressure
(xiii ) Time constraints – when the endoscopy procedure, program, or unit is running significantly
behind schedule, and the training episode is likely to be contributing to this.
Level of agreement 96 %.
Comment Safe and complete LNPCP resection will increase the duration of a procedure. Even
if the LNPCP resection is the only planned activity, adequate time is necessary to
both enable the resection to be completed and to deal with any associated problems
or complications. The allocation of time for procedures can be complex, and will be
influenced by the experience of the endoscopist, the other factors listed above, and
whether the case is being used for training purposes. Failure of the system to ensure
the patent arrives promptly, with the correct bowel preparation administered, and
that the correct time has been allocated will have an impact on the procedure. Time
pressure is perceived differently between individuals; however, time pressure generally
impairs performance because it places constraints on the capacity for thought and
action that limit exploration and increases reliance on well-learned or heuristic
strategies. The resulting time pressure increases speed at the expense of quality.
The outcome of this can be incomplete resection, or a suboptimal technique with a
higher likelihood of recurrence or complications.
4.6 Techniques to manage adverse events related to EMR
29 Management of intraprocedural bleeding
ESGE recommends the best practice technique described below for the management of
intraprocedural bleeding during EMR.
Best practice recommendation, low quality evidence.
(i ) Use the endoscope flushing pump to identify the exact place where bleeding is occurring
within the mucosal defect.
Level of agreement 95 %.
(ii ) Re-injection may help to isolate/make the bleeding vessel visible.
Level of agreement 90 %.
(iii ) Consider the use of image-enhancement techniques (e. g. red dichromatic imaging
[RDI] [135 ]
[136 ]) to identify the bleeding vessel.
Level of agreement 83 %.
(iv ) Ensure the patient position is optimal so that blood flows away from the resection
site to improve visualization (use gravity).
Level of agreement 93 %.
(v ) If significant bleeding occurs, tamponade of the vessel with the endoscope cap or
any instrument may be useful until the required device is available and may help to
identify the source of bleeding.
Level of agreement 95 %.
(vi ) For mild/moderate bleeding, start with STSC [137 ] (e. g. ERBE soft coagulation effect 4, 80 W). The technique is as follows:
use the tip of the snare to apply light pressure precisely over the bleeding point
(confirmed by cessation of bleeding)
apply electrosurgical energy for 1–2 seconds, then lift the snare and re-assess bleeding.
Level of agreement 91 %.
(vii ) If STSC is ineffective after application of the correctly applied technique (as
above), switch instrument to a coagulation forceps.
Level of agreement 87 %.
(viii ) For severe bleeding, start with a coagulation forceps.
Level of agreement 92 %.
(ix ) Use of the coagulation forceps should proceed as follows:
the suspected vessel is captured with the forceps
cessation of bleeding confirms correct forceps placement
the vessel is then gently tented away from the deeper structures and electrosurgical
energy is applied (using the same settings for the electrosurgical unit as for STSC
[e. g. ERBE soft coagulation effect 4, 80 W]) until a visible cautery effect is observed
the tissue is released and the endoscope flushing pump is used to confirm cessation
of bleeding.
Level of agreement 94 %.
(x ) In exceptional circumstances, where thermal energy does not control intraprocedural
bleeding, mechanical control with clips may be required. This may however interfere
with progression of the resection. A clip should be closed on a site of bleeding and
only fired if cessation of bleeding is confirmed.
Level of agreement 91 %.
(xi ) Consider the use of a topical hemostatic agent [138 ] if there is diffuse intraprocedural bleeding that cannot be controlled using the
other methods discussed.
Level of agreement 86 %.
(xii ) The injection of adrenaline (1:10 000) may reduce the rate of bleeding in difficult
circumstances.
Level of agreement 87 %.
30 Management of intraprocedural perforation
ESGE recommends the best practice technique described below for the management of
intraprocedural perforation.
Best practice recommendation, low quality evidence.
(i ) The suspected perforation should be classified according to the Sydney DMI score
[57 ]. DMI types II–V should be closed endoscopically.
Level of agreement 95 %.
(ii ) The endoscopist should first consider completing the resection, or at least clear
the polyp tissue surrounding the perforation, if it is safe/appropriate to do so.
Level of agreement 95 %.
(iii ) In the context of an observed DMI type IV or V with abdominal distension and patient
instability, a 16 G or larger bore needle should be inserted into the peritoneum to
decompress a capnoperitoneum [139 ].
Level of agreement 83 %.
(iv ) The safety and appropriateness of completing the resection (prior to managing the
perforation) depends on: the bowel preparation, competence of the endoscopist, size
of the perforation, stage of the resection, whether peritoneal contamination is imminent,
and patient stability.
Level of agreement 95 %.
(v ) Through-the-scope (TTS) endoscopic clips should be the first-line choice to close
a perforation during EMR.
Level of agreement 95 %.
(vi ) Clip placement should consider the effect of gravity. Start on the same side of
the wound as the fluid pool, so the clip stem falls away from the wound, allowing
further clip placement. Start on an area of normal tissue if possible (not directly
on the perforation) and proceed along the axis of the perforation to achieve a zipper-type
closure.
Level of agreement 96 %.
(vii ) Maximum tissue capture with TTS clipping can be achieved using rotation of the clip
shaft and suction of luminal gas prior to closure.
Level of agreement 95 %.
(viii ) If at this point, closure is not possible consider consulting a more experienced
colleague/trainer if possible.
Level of agreement 89 %.
(ix ) If closure is not possible using TTS endoscopic clips, other techniques available
include: over-the-scope (OTS) clips, nylon poly-loop plus clips, or endoscopic suturing.
These techniques are outside the remit of this competency framework.
Level of agreement 92 %.
(x ) Administer intravenous antibiotics in the case of intraprocedural perforation, considering
any medication allergies/interactions.
Level of agreement 92 %.
(xi ) Post-procedure, consider a computed tomography (CT) scan if there is persistent
pain despite intravenous analgesia following the procedure.
Level of agreement 82 %.
(xii ) Post-procedure, consider review by a surgeon. Base the need for referral to surgery
on the degree of observed fecal contamination, presence/absence of persistent post-procedural
pain, and CT scan appearances.
Level of agreement 92 %.
31 Management of post-EMR bleeding
ESGE recommends the measures given in the list below be considered best practice management
of clinically significant post-EMR bleeding.
Strong recommendation, moderate quality evidence.
(i ) Pre-intervention: resuscitation of patient with fluids and blood products if required.
Level of agreement 96 %.
(ii ) Pre-intervention: address any coagulopathy.
Level of agreement 96 %.
(iii ) Pre-intervention: perform close monitoring after initial resuscitation.
Level of agreement 96 %.
(iv ) Decision for intervention: use the frequency of hematochezia, American Society of
Anesthesiologists (ASA) score, and need for transfusion as a guide to which patients
require repeat colonoscopy and hemostatic measures [140 ]. Most patients do not require re-intervention.
Level of agreement 96 %.
Procedural techniques after a decision to intervene
(v ) Locate the resection defect and use the endoscope flushing pump to irrigate it.
Level of agreement 96 %.
(vi ) There is usually a single large vessel responsible for delayed bleeding after EMR.
This vessel is often identified by an adherent clot. Any clot adherent to the defect
after EMR should be removed using a cold snare and therapy should be directed at the
underlying vessel.
Level of agreement 96 %.
(vii ) Hemostatic forceps should be used following the above technique to cauterize the
causative vessel. The technique is described in Recommendation 29 (ix).
Level of agreement 93 %.
(viii ) Clips can be placed as primary therapy but are often more useful as security after
control is achieved using coagulation forceps.
Level of agreement 95 %.
(ix ) A topical hemostatic agent can be useful, but is often only a temporary measure
to control bleeding.
Level of agreement 90 %.
(x ) In rare cases, where endoscopic therapy fails, consider interventional radiology
[141 ] as the next step. Leaving a clip on/near the causative vessel may allow the radiologist
to visualize the causative vessel more easily in this situation.
Level of agreement 96 %.
(xi ) If endoscopic therapy fails and interventional radiology either fails or is unavailable,
consider surgery.
Level of agreement 96 %.
5 After EMR
32 Alarm features suggesting an adverse event post-EMR
ESGE recommends that the symptoms/signs in the following list be considered alarm
features post-EMR suggesting an adverse event relating to the procedure.
Best practice recommendation, low quality evidence.
(i ) Persistent abdominal pain despite simple analgesics [142 ]
[143 ].
Level of agreement 97 %.
(ii ) Persistent abdominal distension associated with abdominal pain.
Level of agreement 96 %.
(iii ) Hemodynamic instability/shock (fever, tachycardia, hypoxia etc.).
Level of agreement 97 %.
(iv ) Peritonism on clinical examination.
Level of agreement 97 %.
(v ) Persistent large-volume rectal bleeding.
Level of agreement 97 %.
33 Post-procedure discussion with patients
ESGE recommends the topics given in the following list (representing the minimum content
for a competent practitioner to achieve a safe discharge of the patient into the community)
should be covered with a patient in the post-EMR conversation.
Best practice recommendation, low quality evidence.
(i ) Give information about which adverse events could still occur after the procedure
(bleeding, serositis, delayed perforation).
Level of agreement 96 %.
(ii ) The criteria for seeking medical attention should be discussed with the patient.
As a minimum, the patient should be instructed to seek medical attention (perhaps
at the institution where the resection was performed, with specific contact details
given) if they experience: persistent abdominal pain that does not respond to simple
analgesics; frequent hematochezia, with an explicit threshold regarding when to seek
medical attention (perhaps blood in the stool every 30 minutes or less).
Level of agreement 94 %.
(iii ) Give an explanation about post-procedural diet. After standard EMR, this could be
clear fluids overnight and standard diet the day after, but this should be modified
depending on the complexity and radicality of the procedure.
Level of agreement 87 %.
(iv ) Individualized advice should be given about the timing of restarting antiplatelet/anticoagulant
medication based on the published evidence and in consultation with the prescribing
specialists.
Level of agreement 96 %.
(v ) A follow-up plan to discuss pathology results and potential outcomes should be made.
Discuss the expected time until histopathology results are known.
Level of agreement 96 %.
(vi ) The need for surveillance colonoscopy based on the size and histology should be
estimated. It should be explained that this will be confirmed once the histopathology
results are known.
Level of agreement 93 %.
(vii ) If the colonoscopy identifies malignancy, an LNPCP in a young patient, or a suspicion
of a hereditary syndrome, screening of family members following established guidelines
should be recommended [144 ].
Level of agreement 93 %.
(viii ) If the procedure was not completed, the reasons why and the next steps in management
should be discussed.
Level of agreement 96 %.
(ix ) The importance of not travelling to remote areas (without access to emergency medical
care) for a minimum of 14 days should be discussed.
Level of agreement 83 %.
34 Situations requiring admission to hospital
ESGE recommends consideration of admission to hospital immediately after EMR in the
situations given in the following list.
Best practice recommendation, low quality evidence.
(i ) A highly co-morbid patient.
Level of agreement 88 %.
(ii ) A patient who is elderly and/or lives alone without support (especially if sedation
has been used).
Level of agreement 93 %.
(iii ) A patient requiring early re-introduction of anticoagulation (e. g. metal mitral
valve).
Level of agreement 83 %.
(iv ) Intraprocedural perforation with observed fecal peritoneal contamination or incomplete
confidence of closure.
Level of agreement 96 %.
(v ) Significant periprocedural bleeding, even if controlled.
Level of agreement 96 %.
(vi ) Abdominal pain that does not resolve with simple analgesics (e. g. intravenous paracetamol).
Level of agreement 95 %.
(vii ) Unexplained fever/tachycardia.
Level of agreement 95 %.
6 Surveillance procedures after EMR
6 Surveillance procedures after EMR
35 Essential features during surveillance colonoscopy
After endoscopic resection, ESGE recommends the features detailed in the following
list should be confirmed and documented during a surveillance colonoscopy.
Strong recommendation, high quality evidence.
(i ) Assessment of quality of bowel preparation.
Level of agreement 95 %.
(ii ) Completeness of the colonoscopy.
Level of agreement 96 %.
(iii ) Recognition of the EMR scar, extensive imaging (high definition, virtual chromoendoscopy,
+/− chromoendoscopy, near-focus/magnification) and photodocumentation [87 ]
[145 ].
Level of agreement 95 %.
(iv ) Presence of any residual or recurrent adenoma recognized by a transition in pit/vascular
pattern (e. g. NICE I to NICE II; Kudo I to Kudo III/IV) ([Fig. 3 ]).
Level of agreement 96 %.
(v ) The presence of clip artifact [146 ]
[147 ]. Recognize that clip artifact itself does not need treatment, but, where there is
doubt, treatment is safe and prevents unnecessary further procedures for patients.
Level of agreement 96 %.
(vi ) Targeted biopsies and endoscopic treatment of any potential residual or recurrent
adenoma should be performed.
Level of agreement 96 %.
(vii ) Additional polyps/LNPCPs should be noted and removed if feasible or a plan for their
removal made.
Level of agreement 99 %.
(viii ) Need/timing for further surveillance [58 ].
Level of agreement 93 %.
Fig. 3 Images showing a standardized approach to scar assessment after endoscopic mucosal
resection.
36 Treatment of LNPCP recurrence at an EMR scar
ESGE recommends the techniques in the following list be considered best practice for
the treatment of LNPCP recurrence at an EMR scar.
Trainees should aspire to develop competency in these techniques to treat adenoma
recurrence.
Best practice recommendation, low quality evidence.
Small (< 5 mm) residual
(i ) Cold or hot snare polypectomy.
Level of agreement 91 %.
(ii ) Cold avulsion (Recommendation 24) followed by STSC [9 ].
Level of agreement 88 %.
(iii ) Injecting a scar containing a small residual often creates a canyon effect and should
therefore be avoided.
Level of agreement 83 %.
Large (≥ 5 mm) residual
(iv ) Standard injection and snare technique. This should be performed resecting a margin
of normal tissue to isolate the central non-lifting component for snare resection.
Consideration should be given to using an underwater technique [148 ].
Level of agreement 89 %.
(v ) Other techniques exist to treat residual or recurrent adenoma at an EMR scar (e. g.
full-thickness resection, endoscopic submucosal dissection, hot avulsion, forced argon
plasma coagulation, and underwater EMR [149 ]). They have not been subjected to systematic controlled study in this context and
are outside the scope of this competency framework (see also Recommendation 24).
Level of agreement 88 %.
7 Completion of training: sign off and initial credentialing
7 Completion of training: sign off and initial credentialing
37 Completion of training in EMR
ESGE recommends the requirements detailed below for completion of training and initial
credentialing in EMR.
Best practice recommendation, low quality evidence.
(i ) At least 30 EMR procedures performed independently over a minimum period of 6 months
are required. Beyond this, achievement of specific competencies (below), rather than
procedure number, should determine whether an individual is deemed competent to perform
EMR.
Level of agreement 83 %.
(ii ) It is not a requirement to achieve competency in all complexities of LNPCP to be
able to perform any EMR.
Credentialling can perhaps provide a guide as to the ceiling of complexity that is
appropriate for a given practitioner of EMR. In the absence of an evidence-based tool
to assess a given EMR, the trainer should give an idea of the potential SMSA range
that the trainee leaving training should attempt independently.
Level of agreement 89 %.
(iii ) Practicing endoscopists should avoid attempting LNPCPs outside their competency
(SMSA score) unless within the context of a training scheme.
Level of agreement 91 %.
(iv ) Conversely, attempting EMR of an LNPCP one threshold above competency (SMSA score)
within a structured training environment may be entirely appropriate to further the
skills of the trainee.
Level of agreement 89 %.
Comment We have included a tool to assess the quality of polypectomy in this document – the
GPAT ([Table 4 ]). The components of the GPAT that are mandatory depend on whether hot or cold snare
polypectomy is being described. The score for each component can be between 1 and
5. Mandatory components are always included in the denominator; non-mandatory components
only contribute to the denominator if scored. The GPAT is calculated alongside the
SMSA and SMSA+ scores but does not include them in the calculation.
The GPAT score is generated as follows:
The GPAT is a work in progress. It will undergo prospective evaluation to ensure that
it performs at least as well as the other instruments available.
Table 4
Components of the Global Polypectomy Assessment Tool (GPAT).
Component
Possible responses and scoring
Mandatory
Recommendation number
Hot snare
Cold snare
Global competencies
Tip control
1
Uncontrolled, shaky, and undirected
x
x
None
5
Controlled, stable and purposeful
Fully appreciates/demonstrates extent of the polyp to be resected
1
Focusses on one area, does not demonstrate appreciation of the entire polyp
x
x
3(i), 3(ii), 3(iii)
5
Clearly appreciates entire extent of the polyp; approach and resection reflect this
Positioning with respect to the polyp
1
Lesion not at 6 o’clock, far from the colonoscope, fluid covering lesion (poor use
of gravity)
x
x
17(iii), 20(iii)
5
Lesion at or near 6 o’clock, close to the colonoscope, fluid lies away from lesion
(good use of gravity)
Technique selected is appropriate for the polyp resected
1
No clear need for en bloc resection if selected, lesion unsuitable for cold snare,
hot snare for polyp < 10 mm
x
x
4 [53 ]
5
Correct decision for en bloc vs. piecemeal resection, hot vs. cold appropriate for
the polyp
Injection technique
Injection is performed in the correct plane
1
Injection infrequently results in sustained submucosal lifting (transmural / intramucosal
injection)
x
16(iv), 16(v), 16(vi), 16(vii)
5
The submucosal plane is quickly found and injection rapidly results in sustained mucosal
lifting (needle in submucosa)
Injection is performed dynamically
1
Once the needle is situated in the submucosa, there is no movement of the needle away
from the muscularis toward the center of the lumen
x
16(iv)
5
Once the needle is in the submucosa, there is graduated movement of the needle away
from the muscularis towards the center of the lumen
Injection is used to improve lesion access
1
Injection does not facilitate access to the target lesion
x
16(i), 16(viii)
5
Injection clearly facilitates access to the target lesion
Snare placement technique
Appropriate snare size/type selected
1
Snare clearly too large/small and of incorrect type (thin wire vs. thick wire) for
the polyp
x
x
17(i), 17(ii)
5
Snare of appropriate size and type for the polyp
Stable position with lesion at 6 o’clock OR transformed to 6 o’clock
1
Snare position is not consistently maintained at 6 o’clock and/or the position is
unstable
x
x
17(iii)
5
Snare position is consistently maintained at 6 o’clock and the position is stable
Maximizing snare capture
1
Poor capture of tissue/scrapes the surface of the polyp/no use of downward pressure/no
use of gas aspiration/may result in incomplete mucosal layer excision
x
x
17(viii)
5
Good capture of polyp tissue within snare/use of downward pressure/use of gas aspiration
resulting in complete capture of adequate target tissue
Snare precisely visualized during placement and closure (V of the snare)
1
Snare V at intersection with snare catheter not visualized during closure and far
from the colonoscope
x
x
17(vii), 17(ix), 17(x)
5
Snare V visualized consistently during closure and near to the colonoscope
Residual tissue islands avoided if piecemeal resection or macroscopically complete
if en bloc resection attempted
1
Snare placement does not include normal margin (at edge) or does not use transected
tissue edge (within lesion) as a guide resulting in tissue islands/incomplete en bloc
resection
x
x
17(xxviii)
5
Snare placement includes > 2–3 mm normal margin (at edge) of tissue or uses transected
tissue edge as a guide (within defect) resulting in no tissue islands/complete en
bloc
Safety checks prior to resection (hot snare only)
Moves the closed snare to confirm independent movement from deeper structures
1
Does not check tissue mobility prior to transection with respect to deeper structures
x
20(xv), 20(xvi)
5
Checks mobility prior to transection with respect to deeper structures
Lifts the snare away from the muscularis propria prior to application of electrosurgical
energy
1
Does not lift the snare prior to applying electrosurgical energy
x
20(xx)
5
Lifts the snare away from the muscularis prior to the application of electrosurgical
energy
Defect assessment after resection
Mucosa : looks for, detects, and removes residual adenomatous tissue at margin and within
defect
1
Does not ostensibly and systematically check for residual adenomatous tissue at the
defect margin or within the defect, and/or does not remove successfully
x
x
22(i), 22(ii), 22(iii), 22(iv)
5
Ostensibly and systematically checks for residual adenomatous tissue within the defect
and at the defect margin, and removes it successfully
Thermal ablation of the post-EMR margin
1
Unsteady application, results in areas of incomplete ablation, ablates visible polyp
tissue, messy result
5
Steady systematic application, does not ablate visible polyp tissue, complete ablation
of the entire margin achieved
Submucosa : looks for, detects, and treats any bleeding vessels within the defect
1
Neither detects nor treats bleeding vessels in submucosa; treats benign submucosal
appearances
x
22(v), 22(vi), 22(vii), 22(viii)
5
Detects and treats bleeding vessels in the submucosa; does not treat other submucosal
appearances including herniating vessels
Muscularis : looks for, detects, and treats deep mural injury ≥ II (Sydney classification)
1
Misses signs of deep mural injury (types II–V) which require clip closure
x
22(ix), 22(x), 22(xi)
5
Detects and treats types II–V deep mural injury or confirms they are not present
Accessory techniques in polypectomy
Placement of through- the-scope clips
1
Poor tissue capture, poor use of suction and positioning to maximize correct orientation
and amount of tissue captured
29(x)
5
Good use of suction, positioning, and rotation to capture required tissue and achieves
secure appearing closure
Use of polyp retrieval device
1
Poor positioning, does not capture all pieces, does not use sequential place and retrieve
technique
25(i)
5
6 o’clock position, sequential place and retrieve technique applied, captures all
pieces successfully
Use of coagulation grasper
1
Does not use water, does not wait for cessation of bleeding after forceps closure
prior to application of electrosurgical energy, does not tent vessel away from the
muscularis to apply electrosurgical energy
29(ix)
5
Uses water to identify the causative vessel, confirms correct placement with cessation
of bleeding after closure, tents vessel away from the muscularis to apply electrosurgical
energy
Key benefits of the GPAT include:
GPAT is associated with an online logbook that can be used to track progress in polypectomy.
The online tool can be accessed at: www.gieqs.com/gpat-form .
7.1 Priorities for further research into credentialing and potential future statements
to test in a research setting
The SMSA score is the best evidence-based tool available to assess the complexity
of EMR; however, it does not assess the quality of a given polypectomy. Nor do other
existing systems to assess polypectomy (e. g. DOPyS [150 ]) either: (i) allow objective assessment with good interobserver agreement amongst
raters; (ii) provide the ability to assess and deconstruct a video after the event;
(iii) provide an online portfolio to collect data and report outcomes; or (iv) provide
an objective assessment of technique as per the latest evidence in this competency
framework.
Therefore, we believe the future of credentialing lies in a tool that can reliably
assess technique amongst raters. The GPAT was developed for this purpose. We have
included research priorities based upon it below. We envisage updating this section,
as required, as people start to use the score.
There are two potential future points for Recommendation 37. (v ) A single specific GPAT threshold (yet to be determined) attained over a minimum
of 30 procedures (considering only the last 3 months of data) should be obtained at
the end of training and allows credentialling for all difficulties of EMR.
(vi ) After a period of training where a specific GPAT threshold is obtained, a further
training period may be undertaken (minimum 3 months, 30 procedures) in an attempt
to augment the attained threshold. This should only be undertaken in the context of
a training scheme ([Table 5 ]).
Table 5
Possible conversion of GPAT thresholds to SMSA scores for increasing difficulty of
polypectomy, allowing for potential accreditation of specific difficulties of polypectomy.
Possible GPAT threshold for accreditation[1 ]
SMSA score
0.13 (0–0.25)
2
0.38 (0.26–0.5)
3
0.63 (0.51–0.75)
4
0.88 (0.76–1.0)
+
GPAT, Global Polypectomy Assessment Tool; SMSA, size, morphology, site, and access.
1 These numbers need further validation in large-scale prospective study. Minimum of
30 procedures required.
8 Lifelong key performance indicators (quality assurance)
8 Lifelong key performance indicators (quality assurance)
The following recommendations can also be used to assess the quality of an EMR service
and should be available to patients should they request them. An online tool to prospectively
collect these indicators is in development and will be attached here once it is available.
38 Quality indicators of EMR
Essential quality indicators of EMR that should be prospectively monitored, recorded,
and acted upon if deficiencies are identified are those detailed in [Table 6 ].
Strong recommendation, low quality evidence.
Table 6
Essential quality indicators of EMR that should be prospectively monitored, recorded,
and acted upon if deficiencies are identified.
Recommendation 38
Origin
Origin data, % (95 %CI)
Desired standard[1 ]
Minimum standard
Consensus[2 ]
(i) Practitioners of EMR should be aware of their own outcomes and follow them up
Delphi
–
–
–
90 %
(ii) Number of procedures performed independently per year (lesions ≥ 20 mm)
Delphi
–
70 procedures
–
86 %
(iii) Pre-procedure: Clear documentation of patient consent discussion re: risks,
benefits, and alternatives (see recommendation 10)
Delphi
–
95 %
90 %
90 %
(iv) LNPCP attempted for EMR (vs. those deemed endoscopically unresectable for technical
reasons or due to risk of submucosal invasion)
iACE
94.6 %
95 %
90 %
82 %
(v) LNPCP successfully resected in a single session using EMR
iACE
91.4 %
91 %
86 %
89 %
(vi) Rate of intraprocedural bleeding
iACE
13.6 %
14 %
19 %
83 %
(vii) Rate of intraprocedural perforation
iACE
3.3 %
3.5 %
5 %
85 %
(viii) Rate of complete perforation closure
Delphi
–
95 %
90 %
88 %
(ix) Rate of clinically significant post-EMR bleeding
iACE
5.9 %
6.0 %
8.5 %
88 %
(x) Rate of delayed perforation
iACE
0.7 %
1.0 %
1.5 %
89 %
(xi) Rate of unplanned hospitalization/readmission
iACE
3.0 %
3.0 %
4.5 %
89 %
(xii) Rate of surgery for incomplete EMR, adverse event, or malignant histology (patient
does not enter surveillance)
iACE
4.1 %
4.1 %
6.2 %
81 %
(xiii) Patients returning for surveillance examinations
Delphi
–
90 %
80 %
90 %
(xiv) Rate of adenoma recurrence at first surveillance (endoscopically determined),
overall
iACE
13.9 %
14 %
16 %
80 %
(xv) Rate of adenoma recurrence at first surveillance (endoscopically determined),
if complete thermal ablation used at index procedure [38 ]
iACE
1.4 %
1.5 %
2.6 %
89 %
(xvi) Rate of adenoma recurrence able to be treated endoscopically
iACE
96.0 %
96 %
92 %
88 %
(xvii) Rate of surgery for inability to resect recurrence
iACE
1.2 %
1.2 %
4.4 %
79 %
(xviii) Patient satisfaction recorded including comfort scores
Delphi
–
90 %
80 %
81 %
EMR, endoscopic mucosal resection; iACE, international ACE study[3 ].
1 Desired standard is based on the published expert rate (iACE) data or the Delphi
consensus if not available; it refers to the minimum standard or a modification of
the iACE standard agreed by Delphi consensus.
2 Consensus refers to the agreement on the described statement during the voting rounds
that led to this position statement.
3 The international ACE study refers to a prospective database of > 5000 EMR procedures,
which is the largest study in colonic EMR [4 ]; it is based in Australia but now includes four international centers. Data are
correct as of 18 October 2021.
Disclaimer
ESGE Position Statements represent a consensus of best practice based on the available
evidence at the time of preparation. This is NOT a guideline but a proposal for training
in EMR. The statements may not apply in all situations and should be interpreted in
the light of specific clinical situations and resource availability. Further controlled
clinical studies may be needed to clarify aspects of these statements, and revision
may be necessary as new data appear. Clinical considerations may justify a course
of action at variance with these recommendations. This ESGE Position Statement is
intended to be an educational device to provide information that may assist endoscopists
in providing care to patients. The recommendations are not rules and should not be
construed as establishing a legal standard of care or as encouraging, advocating,
requiring, or discouraging any particular treatment. The legal disclaimer for ESGE
guidelines applies to the present position statement.
