Canadian Association of Gastroenterology consensus guidelines on safety and quality indicators in endoscopy

David Armstrong MB BChir1, Alan Barkun MDCM2, Ron Bridges MD3, Rose Carter LLB (Hons)4, Chris de Gara MB MS5, Catherine Dubé MD3, Robert Enns MD6, Roger Hollingworth MD7, Donald MacIntosh MD8, Mark Borgaonkar MD9, Sylviane Forget MD10, Grigorios Leontiadis MD1, Jonathan Meddings MD11, Peter Cotton MD12, Ernst J Kuipers MD13, Roland Valori MD14; on behalf of the Canadian Association of Gastroenterology Safety and Quality Indicators in Endoscopy Consensus Group

Guidelines or position statements on various aspects of quality indicators, safety indicators and credentialing for endoscopy have been developed by the British Society of Gastroenterology (BSG) (1), the American Society for Gastrointestinal Endoscopy and the American College of Gastroenterology (2,4), the Organisation Mondiale d'Endoscopie Digestive/World Organisation of Digestive Endoscopy (OMGE/OMED) (3), the International Agency for Research in Cancer (9) and the Canadian Association of Gastroenterology (CAG) (5)(6)(7)(8). However, these guidelines are all procedure-based and, while they are very important, they either do not address patient needs or do not provide a framework for adoption in the overall context of endoscopy services.
The Global Rating Scale (GRS) was developed from the patients' perspective as a quality improvement tool relevant to the entirety of endoscopy service delivery (10); it was based on quality and safety indicators developed by the BSG (1) and was designed for the English National Health Service in the context of a nascent colorectal cancer screening program. The GRS has been credited with a marked improvement in the quality and availability of endoscopy services in England, and has been adopted in other parts of the United Kingdom. However, although the general principles of the GRS are broadly applicable, specific elements of the GRS and the associated BSG quality and safety indicators are not, necessarily, directly applicable to other health care systems. The development and implementation of endoscopy quality guidelines relevant to different jurisdictions should be based on the needs of the patients and the realities of clinical practice evaluated by providers and users familiar with the relevant health care environment.

AIMS
The primary aim in developing the present consensus guidelines was to identify processes and indicators relevant to the provision of highquality endoscopy services and to achieve consensus on broadly applicable standards and key indicators that will support continuing quality improvement for endoscopy services across many jurisdictions (1,10). The literature review and guideline development addressed only the most commonly performed endoscopic procedures: esophagogastroduodenoscopy and colonoscopy. However, the principal issues addressed in this process are also applicable to more specialized endoscopic procedures, although specific indicators relevant to these procedures fall beyond of the scope of the current project.

Initial stages and identification of the consensus group
The guideline development process was performed in accordance with the Appraisal of Guidelines, REsearch and Evaluation (AGREE) guidelines ( Figure 1) (11).
The goals of the consensus process were defined by the steering committee (DA, AB, RB, RC, CD, RE, CG, RH and DM) and endorsed by the CAG. The voting participants, invited by the steering committee, were selected to ensure multidisciplinary expert input from gastroenterologists, surgeons, nurses and administrators, as well as individuals with expertise in endoscopy, colorectal cancer screening, quality and patient safety (Appendix); two participants (MB and SF) joined the steering group as leads to two of the working groups. Four international experts (BP, PC, EK and RV) and a nonvoting moderator (JM) were also invited. No patients participated; however, consensus participants were provided with a summary of focus-group sessions conducted in a multicentre, qualitative research study that identified patient-derived quality indicators for endoscopy services.
These searches identified 2475 articles. The citations and abstracts were assigned in batches to pairs of assessors from the steering committee; a dedicated, purpose-designed, online voting system was used to identify relevant articles on the basis of the article title and abstract. All conflicts were resolved by consensus. After two rounds of reviewing, 817 references were retained and the complete articles were uploaded to an online literature review site, used by the steering committee to develop draft statements. Additional literature searches were conducted during the statement development process, as needed.

Development of statements and iterative voting process
Fifty initial statements were developed, based on the literature review, analysis of the GRS and input from the steering committee; these were intended to be broad in nature rather than highly specific and they were used as a framework for continued development. An online system enabled the 35 voting participants (Table 1) to engage in three rounds of votes during the development process. At each voting round, participants voted on the importance and content of all statements using a 6-point scale ('disagree strongly', 'disagree with major reservation', 'disagree with minor reservation', 'agree with major reservation', 'agree with minor reservation' and 'agree strongly') and commented on the wording and validity of the statements. Results were compiled by the CAG to ensure voter anonymity. Round 1 vote: Using the 6-point scale, participants rated the extent to which they agreed that each of the 50 initial statements was important and relevant to the quality or safety of endoscopy. After reviewing the results of this round of voting, the steering committee reduced the number of statements to 22 and made iterative changes to the remaining statements to incorporate participants' comments. Round 2 vote: Participants voted to indicate the extent of their agreement with each of the 22 statements using the 6-point scale. After this round of voting, the 22 statements were assigned to nine working groups for further refinement and development of the evidence base before a third round of voting. The nine working groups were convened to address the following topics: quality indicators, safety indicators, endoscopy reporting standards, ethics, credentialing, quality assurance legislation, GRS development, patient perceptions of endoscopy quality and guideline dissemination. The conclusions of the working groups were then reported back to the steering committee. Round 3 vote: Participants again voted to indicate the extent of their agreement with each of the statements, using the 6-point scale. After the third round of voting, the statements were revised and two additional statements were added. The resulting 24 statements were reviewed by the working groups, who allocated relevant references to each statement to enable evidence grading.

Grading of evidence
Evidence grading was performed by two independent evaluators (GL and KK) using a modified Grading of Recommendation Assessment, Development and Evaluation (GRADE) process (Table 2) (12). This process involved two steps: assessment of the methodological quality of individual studies and assessment of the overall level of evidence behind each statement. Each step was performed independently by the two evaluators and disagreements were resolved by consensus.
The overall level of evidence across studies for each statement was assessed using a Summary of Findings table (13). The GRADE system classifies the quality of evidence into one of four levels -high, moderate, low or very low. Evidence based on randomized controlled trials begins as high-quality evidence, but may be decreased due to study limitations (identified through the risk of bias evaluation), inconsistency (heterogeneity), indirectness, imprecision or other considerations such as reporting bias. Evidence based on case-control or cohort studies starts as low-quality evidence, but may be further decreased (for the same reasons as for randomized controlled trials) or increased if the magnitude of the treatment effect is very large, if there is evidence of a dose-response relationship or if all plausible biases would decrease the magnitude of an apparent treatment effect (12). The grading was subsequently reviewed and approved by all consensus members.

Consensus meeting and final voting
At the three-day consensus meeting, held in June 2010, each section was introduced by a member of the relevant working group who summarized the relevant literature and key issues. Each statement was subsequently discussed by the consensus group under the direction of the nonvoting moderator. A statement about withdrawal of consent was incorporated into a more general statement on consent (statement 1). Thus, the final vote included 23 statements.
Following discussion, all voting participants used electronic keypads to record two separate anonymous votes on each statement. The first vote asked whether participants believed the statement should be implemented, and the second asked the extent to which they agreed with the overall recommendation for implementation ( Table 2). In voting on the strength of recommendation, participants were asked to consider desirable and undesirable effects, patient values and preferences, the quality of the evidence and the judicious use of resources (12).
After the meeting, voting participants used the online system to identify specific, individual quality indicators, safety indicators and mandatory endoscopy reporting fields that they would recommend, again using a six-point scale for each item: 'disagree strongly', 'disagree moderately', 'disagree slightly', 'agree slightly', 'agree moderately' and 'agree strongly'.
For each vote, a recommendation was considered to have been adopted by consensus if 80% or more of participants selected 'agree slightly', 'agree moderately' or 'agree strongly'.

Ethics and financial support
The consensus process and meeting were administered by the CAG, supported by Canadian Partnership Against Cancer/Partenariat Canadien Contre le Cancer and Canadian Institutes of Health Research/ Instituts de Recherche en Santé du Canada: Institute of Nutrition, Metabolism, and Diabetes. All members provided conflict of interest statements before the meeting. Honoraria for participation were provided for international experts and participants who were not members of the CAG. Travel and accommodation were provided for all participants. After review of all disclosures, it was determined that there should be no limitations placed on any participant (14).

ConSEnSuS StAtEMEntS SECtIon 1: EthICS Introduction
Informed consent must be obtained from a patient before any endoscopic procedure is performed. This statement provides the legal and ethical background to this position.

Statement 1. For a patient to give a physician informed consent
to perform an elective endoscopic procedure, the patient must be advised, in a timely fashion, of all relevant information about the procedure, its risks, benefits and alternatives, if any, and be given an opportunity to ask questions that the physician must answer.

Discussion
This statement describes the minimum required to document acceptable provision of informed consent, consistent with Canadian law. It is necessary to obtain informed consent in a way that satisfies the patient's expectation of full, informed consent based on individual needs for information to address personal concerns. The patient can be provided with information or instructions (eg, for bowel preparation) in several formats, including videotapes (15)(16)(17), brochures, information sessions led by allied health care professionals (nurses or assistants) or computer-assisted instructions/interactive software (18). However, these forms of communication do not obviate the need for a conversation with the endoscopist; it is the endoscopist who should obtain informed consent. Trainees may obtain consent if they are to perform the procedure and if disclosure is complete. However, in one study, up to one-third of trainees obtaining consent for colonoscopy had not disclosed the risks of perforation and hemorrhage (19). Ultimately, it is the supervisor who is responsible for ensuring that complete information is provided and that questions have been appropriately answered.
For an elective procedure, disclosure should allow the patient sufficient time to understand the information, reflect on it, ask questions and decide whether to proceed or to select an alternative. In an emergency situation, informed consent should be obtained as soon as possible. For direct-access endoscopy, the patient may be sent information before the procedure, although, in one study, 37% of patients would have preferred meeting a gastroenterologist before the colonoscopy and 20% said that they received the most useful information after taking the bowel preparation (20). In all cases, full verbal disclosure, documented in the patient record, is still required before the procedure (Box 1).

box 1 Full disclosure for informed consent:
Required elements for review The choice of sedation (or no sedation) should be a shared decision, made on a case-by-case basis, between the attending physician/ endoscopist and patient, based on the patient's expectations of comfort, the patient's medical condition, and complexity and duration of the procedure. Regardless of the type of sedation recommended, patients should understand that they may decline sedation.
The consent process should account for the patient's competence and understanding. Language, cognitive ability, severity of illness, pain and analgesia may affect a patient's capacity to understand and consent to a procedure. Provision of informed consent is not a guarantee that the patient has understood the information provided; recall of informed consent is often suboptimal. In one study, 50% of patients could not recall important information on their procedure's indication or risks or when consent was obtained (26). Effort should be made to ensure that the patient has the capacity to consent and to maximize their understanding of the information provided, possibly by complementing oral with written information.
Consent must be given without coercion; this implies that consent can be withdrawn at any time. However, it is possible that a patient under sedation may not be fully cognizant of their request to halt a procedure and the associated risks; nonetheless, if a patient requests that the procedure be stopped, the endoscopist should review his or her technique, the patient's comfort and the need to continue. Procedure cessation may put the patient at risk of increased complications and may not be appropriate.

SECtIon 2: FACIlIty StAnDARDS AnD PolICIES
Introduction Endoscopy facilities must meet or exceed defined standards of quality for operating standards, appropriateness of procedure, and technical and personnel resources. Facilities also must have policies in place that ensure best practice before, during and after endoscopic procedures. Statement 2. Endoscopy facilities should meet or exceed defined operating standards, in all domains, consistent with accreditation under the appropriate national or regional standards.

Discussion
Operating standards may be endorsed locally by many endoscopists and endoscopy facilities, but adoption of appropriate standards is not universal and approved; accepted national guidelines are unavailable in many jurisdictions. Although endoscopic facilities differ greatly with respect to structure, size and procedural case loads, operating standards should apply uniformly, regardless of the facility's size, design or location. It is, however, recognized that standards may be achieved by different means in different facilities.
In many jurisdictions, accrediting bodies do not have specific directives for endoscopy facilities. If endoscopy facilities are to be included in hospital accreditation programs, comprehensive directives must be specified to ensure that high-quality care becomes the standard. Ideally, accreditation of hospital and out-of-hospital endoscopy facilities should be governed by comparable processes to support continuous quality improvement programs, with regular review in all cases.
The adoption of standards and procedures to ensure safety and quality should also support appropriate resource use; indicators of appropriate utilization include no-show rates, adequacy of preparation, nurse-to-patient ratios, safe working environment and appropriate flow of patients. Quality assurance activities should be defined from a patient perspective, and should target indicators of procedural completeness and accuracy, opportunities for patient feedback, and evidence of continued change and reassessment to demonstrate improvement (see section 3 below). Quality and safety standards should also address endoscope reprocessing, conscious sedation, monitoring protocols and resuscitation equipment. Although documented transmission of infection from endoscopic equipment is extremely rare (27), the United States Food and Drug Administration recommends that all institutions have a written program and appropriate administrative organization for endoscope processing, handling and storage (28). Every facility providing endoscopic services should ensure the ongoing availability of protocols, equipment, and trained, competent personnel necessary for the provision of safe and effective conscious sedation. level of agreement with recommendation: Agree, 97% (agree strongly, 85%; agree moderately, 6%; agree slightly, 6%; disagree slightly, 3%)

Discussion
The indication for every procedure should be documented in the procedure report, and the indication should be consistent with accepted guidelines. National or regional guidelines are preferable; however, if none are available or applicable, locally relevant guidelines should be developed.
Consensus guidelines provide explicit statements of appropriate indications for endoscopic procedures (29,30), and there is evidence that the diagnostic yield of endoscopy is significantly increased if procedures are performed for appropriate indications (31,32). Regrettably, reports indicate that 11% to 39% of endoscopic procedures are performed for inappropriate indications (31,33), and that surveillance endoscopies may be performed at inappropriate intervals (34) or unnecessarily (35)(36)(37). However, guidelines should not be the only determinants of procedural appropriateness; for example, esophagastroduodenoscopy may be an inappropriate initial step in dyspepsia management, but it may become appropriate if initial medical therapy fails (29). Similarly, a colonoscopic surveillance interval may, appropriately, be shorter than recommended if, for example, bowel preparation is poor. However, in all cases, the reason for deviation from guidelines or accepted indications should be clearly documented. level of agreement with recommendation: Agree, 100% (agree strongly, 69%; agree moderately, 26%; agree slightly, 6%)

Discussion
Published, evidence-based guidelines do, on occasion, specify the technical and personnel resources needed for procedures, such as endoscopic reprocessing and urgent endoscopy, in the event of a major gastrointestinal bleed (28,38). The importance of experienced ancillary personnel for procedural quality and efficiency is highlighted, for example, by reports of higher polyp detections rates at screening colonoscopy in the presence of more experienced nurses (detection rate of 46% with more than six months of experience versus 40% with six months or less; OR 1.26 [95% CI 1.09 to 1.46]) (39) or of efficiency gains of up to 30% when additional personnel were available to gain intravenous access and administer medications (40).
A skills mix review (41) can identify the personnel resources necessary for a particular institution or facility, based on types of procedures performed, types of sedation offered and need for out-of-hours emergency services. level of agreement with recommendation: Agree, 97% (agree strongly, 86%; agree moderately, 9%; agree slightly, 3%; disagree slightly, 3%)

Discussion
Preprocedure assessment should encompass the risks of sedation, the preparation and the procedure, including documented assessments of patient age, comorbidity, American Society of Anesthesiologists classification, significant medical issues, medications (including type of bowel preparation) and potential complications of bowel preparation (such as dehydration). Relevant policies and guidelines should be implemented (Box 2).

box 2 Preprocedure guidelines that should be implemented
• Antibiotic prophylaxis guidelines for prevention of infective endocarditis; eg, American Heart Association (42) and BSG guidelines (43).
• Antithrombotic agent guidelines on management of antithrombotic agents for endoscopic procedures; eg, American Society for Gastrointestinal Endoscopy guidelines (44) state that anticoagulant agents (eg, warfarin, unfractionated heparin, low-molecular-weight heparin) and antiplatelet agents (eg, acetylsalicylic acid, clopidogrel, ticlopidine, glycoprotein IIb/ IIIa receptor inhibitors) increase the risks of procedure-related bleeding. Preprocedural assessment should encompass the benefits, risks and urgency of the procedure, the bleeding risks associated with antithrombotic therapy and the procedure, and the thromboembolic risk of stopping therapy.

Discussion
Intraprocedural policies are important to achieve and maintain maximal patient safety and comfort and optimal examination quality (Box 3). Completion of colonoscopy, with complete cecal visualization, is essential for detecting lesions in the proximal colon. Endoscopists with a higher cecal intubation rate have a higher rate of colonic neoplastic lesion detection (49). Documentation of cecal intubation using still image or, preferably, video recording (50) of the appendiceal orifice and ileocecal valve, should be confirmed by the endoscopy assistant/ nurse. Terminal ileum biopsy should be performed only if indicated clinically. Right iliac fossa transillumination and finger indentation are inadequate to confirm cecal intubation (51). Cecal intubation rates greater than 90%, documented in several case series (52)(53)(54), should be standard of practice for symptomatic patients and intubation rates should exceed 95% for screening colonoscopies (1,3). Conscious sedation (intravenous benzodiazepines, with or without narcotics, or propofol) may improve endoscopy completion rates (55) but is resource intensive. An evidence-based sedation protocol can improve practice quality and reduce sedation-related adverse events (56,57).
Poor bowel preparation is associated with longer procedure time (58), greater patient discomfort (58), lower colonoscopy completion rates (55), and higher risks of adverse events and missed lesions. A standardized tool such as the Ottawa Bowel Preparation scale (59) or the Boston Bowel Preparation scale (60) should be used to assess bowel preparation quality and process changes should be implemented if inadequate bowel preparation is prevalent.
Endoscopy facilities should institute a policy to record the colonoscopy withdrawal time as a surrogate for a thorough examination. Longer withdrawal times are associated with greater detection rates for advanced neoplasia (adenomata of 10 mm or larger, lesions with villous change, high-grade dysplasia or cancer) (61) and advanced neoplasia detection is improved by a protocol specifying a minimum withdrawal time of 8 min (62). level of agreement with recommendation: Agree, 100% (agree strongly, 89%; agree moderately, 9%; agree slightly, 3%)

Discussion
Clear specific discharge policies (Box 4), should be implemented and should accommodate differences in patients' responses, particularly, to sedation. Elderly patients (70 years of age and older) and those with oxygen desaturation, hypotension, bradycardia and those needing reversal agents require longer recovery times (63). A list of criteria such as the Aldrete score (respiration, oxygen saturation, consciousness, circulation and activity levels) should be used to determine readiness for discharge (64,65). The half-life of reversal agents tends to be shorter than the half-life of sedatives; therefore, reversal agent effects may wane before those of the sedative. Discharge policies should specify patient transportation requirements and activity restrictions after the procedure, with respect to the type and dose of sedation.
Patient satisfaction with discharge processes should be assessed regularly, for example, using standardized postdischarge surveys. level of agreement with recommendation: Agree, 90% (agree strongly, 39%; agree moderately, 42%; agree slightly, 9%; disagree slightly, 6%; disagree moderately, 3%) Statement 9. All patients, on discharge, are given written information regarding the procedural findings, plans for treatment and follow-up, worrisome symptoms to watch for, and steps to be taken.

Discussion
On discharge, written details of the procedure should be provided for the information of the patient and of any physician providing subsequent care (eg, for complications). Provision of a procedure report reduced patients' postprocedure anxiety, improved their recall of findings and recommendations and improved adherence to recommendations (66).
The facility's policies should specify the information to be provided in the discharge report (Box 5), the details of follow-up arrangements that have been or will be made and the person responsible for arranging the follow-up plan.

SECtIon 3: QuAlIty ASSuRAnCE Introduction
Maintaining and enhancing the quality and safety of endoscopy services should be a continuous process that measures aspects of endoscopic performance, implements changes based on these measurements, monitors the effect of these changes and evaluates these effects to achieve new standards (Figure 2).

Statement 10. Endoscopy facilities should maintain a comprehensive quality improvement program incorporating formal, regular, scheduled review of performance reports.
Evidence grade: Low/very low Strength of recommendation: Do it, 85%; possibly do it, 9%; possibly don't do it, 3%; don't do it, 3% level of agreement with recommendation: Agree, 94% (agree strongly, 76%; agree moderately, 12%; agree slightly, 6%; disagree slightly, 6%) Statement 11. Endoscopy facilities should appoint a review committee to monitor and report back to management on adherence to and implementation of quality standards.

Discussion
In some jurisdictions, quality assurance activities are legislated and, in many jurisdictions, a quality assurance committee's proceedings are protected from disclosure if it is constituted according to applicable regional, territorial, state or national legislation and regulations (66-77). Thus, endoscopy quality assurance committees must be developed within the appropriate legislative framework to oversee the performance of the endoscopy facility and all of its personnel. The potential conflicts arising from the disparate perspectives of different stakeholders can be resolved if quality in endoscopy is addressed from the perspective of the patient. In developing the GRS for evaluation of endoscopy, two patient-centred quality dimensions were devised (10): the "Clinical Quality and Safety" dimension addressed six items related to appropriateness of the procedure, information and consent, safety, patient comfort, quality and prompt communication of results, while the "Quality of Patient Experience" (customer care) dimension addressed six items related to equality, timeliness, patient choice, privacy and dignity, aftercare and the ability to provide feedback to the endoscopy service.
A quality improvement and assessment tool such as the GRS provides a framework to support evaluation of service delivery by identifying specific indicators, consistent with high-quality endoscopy that should be auditable (ie, relevant outcomes for which there are no set standards) or measurable (ie, relevant outcomes for which there is an established standard).

Statement 12. Endoscopy facilities should systematically and regularly review current indicators of quality for all endoscopic procedures and implement appropriate responses.
Evidence grade: Low/very low level of agreement with recommendation: Agree, 100% (agree strongly, 82%; agree moderately, 15%; agree slightly, 3%)

Discussion
Quality indicators and safety indicators for endoscopy were considered sufficiently distinct that two working groups were charged with identifying measurable or auditable outcomes that would support quality improvement. After the consensus meeting, participants voted on their endorsement of each quality and safety indicator identified by the working groups (Boxes 6 and 7).

Consensus, %
Indicator related to entire endoscopy facility Quality and safety indicators should be recorded systematically; the American College of Surgeons National Surgical Quality Improvement Program (78) is a potential model for a safety monitoring system, although it is labour intensive and would require modification for use in endoscopy.

box 7 Indicators of safety compromise in endoscopy
Indicator Comment Adverse events should be reviewed by the facility's endoscopy quality review committee to determine the indication for the procedure and decision as to the level of attribution (eg, definitely, probably, possibly or unlikely to be due to the procedure), as recommended for good clinical practice in clinical research (79), the severity of the event and the appropriateness of the procedure. Changes should be implemented as appropriate and the impact of changes made should be reviewed within an acceptable time frame (such as three to six months).

Indicators of increased risk of complications
Indicators of safety compromise may be monitored by phone call follow-up, questionnaire mail back or review of hospital admissions. At a minimum, relevant clinical events and near misses should be recorded, systematically, through a safety learning report system to identify risks, reduce the likelihood of complications, improve safety of the service and reassure patients and physicians.

SECtIon 4: tRAInInG, EDuCAtIon, CoMPEtEnCy AnD PRIvIlEGES Introduction
To provide high-quality endoscopy services, endoscopy staff's skills must meet predefined standards. This requires that endoscopy facilities provide staff with the opportunities to evaluate, maintain and improve their skills on a regular basis while accommodating the disparate needs of patients, trainees, trainers and the facility. level of agreement with recommendation: Agree, 100% (agree strongly, 66%; agree moderately, 26%; agree slightly, 9%)

Discussion
Endoscopy facilities should provide continuing training and competency assessment for all staff in all their areas of activities, including clinical policies, patient monitoring, administration of intravenous medications, endoscope handling and maintenance, use of accessories, emergency procedures, decontamination, patient recovery, communication skills and new technologies. The benefits of nurse training, for example, include improved polyp detection rates (39) and nonpharmacological management of pain and anxiety. This statement does not imply that the facility should pay for all education opportunities or support 'growth opportunities' that are not congruent with the facility's needs. It does, however, imply that all staff members have the opportunity to evaluate and maintain skills relevant to their position. level of agreement with recommendation: Agree, 100% (agree strongly, 94%; agree moderately, 3%; agree slightly, 3%) Statement 16. All endoscopy facility personnel engaged, directly or indirectly, in endoscopy service delivery should be trained and certified as having competency to perform specified routine and/ or emergency procedures according to appropriate current standards.

Discussion
Competency is the minimal level of skill, knowledge and expertise derived through training and experience required to perform a task or procedure safely and proficiently, without assistance or supervision (80). Endoscopists must be competent to perform endoscopic procedures safely and to interpret and manage findings correctly. Competency for specific procedures should be assessed by objective measures with full documentation of the number and type of procedures performed. Methods for determining competency should include guidelines on the role of observers, number of cases observed and criteria assessed. Competency assessment should encompass proficiency in common therapeutic interventions, cognitive competency (ie, how to manage the patient overall) and technical competency.
Minimal levels of endoscopic competency should be defined, consistent with the OMGE/OMED guidelines on credentialing and quality assurance in digestive endoscopy (Box 8) (3). The minimum numbers of procedures recommended by many professional associations as thresholds for evaluation of competency, may be markedly lower than numbers required for documentation of proficiency (81,82).

box 8 Competencies required by end of training (oMGe/oMeD) (3)
• Successful completion of a recognized medical or surgical training program • Ability to integrate endoscopy into the clinical management plan (be this medical, surgical or referral for specialty services) • Understanding of indications, contraindications and risks related to procedures • Ability to clearly describe to the patient, in layman's terms, details of the procedure including attendant risks and, thus, to obtain informed consent level of agreement with recommendation: Agree, 97% (agree strongly, 66%; agree moderately, 29%; agree slightly, 3%; disagree slightly, 3%)

Discussion
Under the auspices of a properly constituted quality assurance committee (see statements 10 and 11) (66-77), endoscopists should review their endoscopy practice data regularly including the number and type of procedures, and standard quality and safety indicator outcomes (eg, completion rates, unplanned events, comfort scores, patient satisfaction). Practice audits (83)(84)(85)(86) to facilitate maintenance of competence should be complemented by a formal annual evaluation of endoscopy performance. Any serious unplanned event or series of events should be reviewed by the endoscopist and quality assurance committee with documentation of deficiencies, discrepancies and any consequent actions. Audits can detect poor performance/inadequate numbers and can improve outcomes such as colonoscopy completion rate (52,(87)(88)(89). Unverified self-reporting is not encouraged as a formal performance indicator, although this does not preclude self-reflection based on structured practice audits directed at personal learning programs. Verifiable databases, generated from electronic reporting systems or web-based practice audits, have greater validity for documenting, but these do require that endoscopy facilities have tools available to monitor outcomes. level of agreement with recommendation: Agree, 100% (agree strongly, 91%; agree moderately, 9%) Statement 19. Endoscopists' privileges should be subject to formal, regular, scheduled review to ensure that renewal is based on documented competence to perform specified procedures consistent with appropriate current standards.

Discussion
Health care institutions are legally responsible for ensuring that individuals who perform procedures are competent (3). Endoscopy privileges are granted or renewed by health care institutions, including hospitals and out-of-hospital endoscopy facilities, to competent individuals based on regional and national guidelines and regulations.
Policies guiding the granting of endoscopic privileges should be applied uniformly for all endoscopists; privileges should be granted separately for each endoscopic procedure, based on competence determined according to current standards (5)(6)(7)(8). Each institution should specify appropriate standards, monitor adherence to standards and update standards, as needed.
When applying for privileges, the applicant and training program director must provide details of training undertaken for each endoscopic procedure requested. Each application should be reviewed by a clinician who is knowledgeable about the relevant procedure; procedural competence should be documented from direct observation of the applicant's performance at the privileging institution or by the training program director.
Endoscopic privileges should be granted for a finite period to permit regular re-evaluation of the applicant's performance and competency; procedural volumes should also be evaluated because, for example, colonoscopic complications have been associated with procedures performed by lower-volume endoscopists (90). Renewal of privileges should be based on defined policies for addressing poor performance.

SECtIon 5: EnDoSCoPy REPoRtInG StAnDARDS Introduction
A completed, comprehensive endoscopy report is an essential element of a quality endoscopy service. Traditional, narrative reporting is associated with marked variations in the documentation of positive findings, pertinent negative findings and other procedural details. Effective communication of procedural findings and successful practice audit and quality improvement processes are virtually impossible in the absence of a standardized, ideally electronic, endoscopy report format. level of agreement with recommendation: Agree, 97% (agree strongly, 76%; agree moderately, 15%; agree slightly, 6%; disagree strongly, 3%)

Discussion
Deficits in endoscopy reporting include marked variability in the definitions of inflammation in ulcerative colitis (91), as well as marked differences in the completion of different report elements such as lesion identification and removal (84% of reports), sedation procedure (75%), demographic data (69%), procedure interpretation (58%), patient history (57%) and procedure quality (40%) (92).
Electronic reporting offers a data collection method that can be more complete and cost-effective -but no more time-consumingthan handwritten reports or free-text dictated reports (93). Although electronic reporting is initially more expensive than handwritten or dictated reporting (94), the overall costs become comparable after five years, and cost-benefit outcomes are better after three years, in part, because automatic electronic transmission of reports reduces costs, administrative workload and communication delays.
Standardization of electronic reports, including mandatory reporting elements (Box 9) and accepted grading systems (Box 10), permits standardization of the data capture and long-term storage (at least 10 years) needed for audit and quality improvement processes. While electronic data (eg, digital transcripts of dictated reports) are preferable to handwritten reports, they are inferior to a full, structured electronic report.  (99) • Crohn's disease -SES-CD activity score (100) • Ulcerative colitis disease activity: Mayo score (101,102) • Bleeding lesions -Forrest classification (103) • Esophageal varices -grade size (104,105) • Readiness for discharge: Aldrete score (64,65) Statement 21. Endoscopy facilities should implement policies to monitor and ensure the timeliness and completeness of procedure reporting.
Evidence grade: Low/very low

Discussion
An important aspect of clinicians' competency relates to the timely provision of a completed procedure report. Receipt of a procedure report improves patients' adherence to follow-up appointments and therapies (66). Optimally, the endoscopy report should be available on the day of the endoscopy, either as a modified, patient-centred version or as the formal report, supplemented by a patient-centred summary.
When relevant, pathology results should accompany the final endoscopy report but this can be a technical challenge because it requires linkage of separate databases. Patient concerns regarding their test results should be addressed in a timely manner although the nature and timeliness of the response may, legitimately, vary among institutions, depending on local needs and resources. level of agreement with recommendation: Agree, 100% (agree strongly, 71%; agree moderately, 26%; agree slightly, 3%)

Discussion
Acknowledgement of the patient's perspective on all aspects of endoscopy service delivery and responsiveness to their concerns necessitates a collaborative relationship between the clinician and the patient to address the quality of the patient's experience and the appropriateness, accuracy and safety of the procedure. Patients' perspectives on quality endoscopy care were reviewed at the consensus conference based on the results of focus groups held in French and English Canada for adults who had undergone or were scheduled to undergo a colonoscopy. Focus group participants' discussions of the total colonoscopy experience, defined as all factors and events occurring before, during and after the day of the procedure, yielded several major quality themes, including communication: quantity and quality; comfort: physical and psychological; and attitude and demeanor: physician and endoscopy unit staff. Endoscopist expertise, procedural safety and physical amenities were also important. Many patients expressed little concern about quality indicators (106)(107)(108), reporting that they trusted the accuracy and safety of the test on the presumption that tests were closely monitored and regulated to ensure adherence to standards.
Preprocedural educational interventions, such as provision of additional information, as an educational video (16,109), booklet (110), face-to-face discussion (111) or behavioural intervention (112) can support patients' need to understand the procedure and improve their satisfaction and appointment keeping. Interventions may have a limited effect, due to high preprocedural satisfaction levels (113), but they can, nonetheless, improve patients' knowledge about the procedure.
In theory, patients should be more likely to comply with follow-up examinations if they experienced high levels of satisfaction with their previous endoscopic procedure; to date however, this hypothesis remains unproven.
Statement 23. Endoscopy facilities should systematically and at least annually solicit patient feedback, report the results to the service and to the institution's quality committee, and implement effective measures to address patients' concerns.

Discussion
The increasing and appropriate emphasis on patient-centred care requires that patients provide feedback on their experience to inform endoscopy facilities' decisions on balancing cost and volumes, access and wait times, discomfort and completion rates, and resources and procedural times. Patient feedback should be sought, at least annually (10), using one or more methods including satisfaction surveys, focus groups and invited comments. Patient feedback is essential for assessing many aspects of endoscopy service delivery, such as the quality of the informed consent process (114), and offers measurable improvements in patient experience (115). Regular, structured measurement of patient satisfaction ensures that patients' perspectives are evaluated alongside more traditional indicators such as appropriateness, accuracy and safety.

ConCluSIonS
Gastrointestinal endoscopy is a complex diagnostic and therapeutic undertaking that demands a high level of skill and knowledge on the part of the operator (2,5). However, high-quality endoscopy requires more than a skilled operator -the delivery of high-quality endoscopy services, in a cost-effective manner consistent with the broader needs of a health care system, requires a formal quality improvement framework that addresses all aspects of endoscopy service delivery from the patient's initial contact with a health care provider (eg, the identification of family history of colon cancer in an asymptomatic individual) through to documentation of long-term outcomes (eg, freedom from colon cancer over decades). Recognition of the patient as the focus of the endoscopy process provides a structure for integrating the efforts of the many diverse disciplines whose contribution is needed to ensure a high-quality service.
The fundamental principle underlying high-quality health care is the need for an iterative feedback loop centred on the patient's needs. The feedback loop requires identification of the patient's needs, measurement of the extent to which these needs are met, intervention to ensure that unmet needs are addressed and reassessment of the patient's needs after intervention. The GRS (10) is a quality improvement tool that enables the adoption of iterative quality improvement processes in endoscopy, supported by the identification of quality and safety indicators relevant to specific aspects of service delivery.
Whenever possible, health care recommendations are based on the highest quality evidence available. Outcomes of this consensus conference confirm that there is a paucity of high-grade evidence relevant to endoscopy service delivery. However, among a large multidisciplinary group of health care professionals, there was a high level of agreement on key features that should be addressed to improve endoscopy quality. The fact that the supporting evidence was generally graded as 'low' or 'very low' does not indicate that the recommendations, themselves, are weak; it indicates, rather, the need for concerted, widespread efforts to document the short-and long-term effects of adopting new quality improvement processes in endoscopy.
The consensus process, presented in the current report, provides the framework for a quality improvement structure in endoscopy based on explicit recommendations that will support systematic monitoring, assessment and modifications in endoscopy service delivery; this framework is intended to yield benefits for all patients whose care may be affected by the practice of gastrointestinal endoscopy.
Guarantor of the article: David Armstrong. Specific author contributions: All authors played a role in providing the content for the manuscript (see 'working groups', above). All authors approved the final draft of the manuscript. Financial support: The meeting was supported by Canadian Partnership Against Cancer/Partenariat Canadien Contre le Cancer and Canadian Institutes of Health Research/Instituts de Recherche en Santé du Canada: Institute of Nutrition, Metabolism, and Diabetes. Honoraria for participation were provided to the international experts and for participants who were not members of the CAG. Travel and accommodation was provided for all participants.