The European Society of Gastrointestinal Endoscopy Quality Improvement Initiative: developing performance measures

Abstract

The European Society of Gastrointestinal Endoscopy (ESGE) and United European Gastroenterology (UEG) have a vision to create a thriving community of endoscopy services across Europe, collaborating with each other to provide high quality, safe, accurate, patient-centered and accessible endoscopic care. Whilst the boundaries of what can be achieved by advanced endoscopy are continually expanding, we believe that one of the most fundamental steps to achieving our goal is to raise the quality of everyday endoscopy. The development of robust, consensus- and evidence-based key performance measures is the first step in this vision. ESGE and UEG have identified quality of endoscopy as a major priority. This paper explains the rationale behind the ESGE Quality Improvement Initiative and describes the processes that were followed. We recommend that all units develop mechanisms for audit and feedback of endoscopist and service performance using the ESGE performance measures that will be published in future issues of this journal over the next year. We urge all endoscopists and endoscopy services to prioritize quality and to ensure that these performance measures are implemented and monitored at a local level, so that we can provide the highest possible care for our patients.

Abbreviations

ADR

adenoma resection rate

AGREE

Appraisal of Guidelines for Research and Evaluation

AMSTAR

Assessing the Methodological Quality of Systematic Reviews

ASGE

American Society for Gastrointestinal Endoscopy

CARE

Complete Adenoma Resection [study]

CIR

cecal intubation rate

CRC

colorectal cancer

EOI

expression of interest

ERCP

endoscopic retrograde cholangiopancreatography

ESGE

European Society of Gastrointestinal Endoscopy

GI

gastrointestinal

GRADE

Grading of Recommendations Assessment, Development and Evaluation

ISFU

Importance, Scientific acceptability, Feasibility, and Usability

NQMC

National Quality Measures Clearinghouse

PCCRC

post-colonoscopy colorectal cancer

PICOS

population/patient, intervention, comparison, outcome, study design

QUADAS

Quality Assessment Tool for Diagnostic Accuracy Studies

QIC

Quality Improvement Committee

SIGN

Scottish Intercollegiate Guidelines Network

UEG

United European Gastroenterology

The importance of quality

Tens of millions of people undergo endoscopic procedures every year in Europe. Endoscopy is the pivotal investigation in the diagnosis of gastrointestinal pathology and a powerful tool in its management. High quality endoscopy delivers better health outcomes and a better patient experience.¹ yet there is clinically significant variation in the quality of endoscopy currently delivered in endoscopy units.^2–6

An example of this is post-colonoscopy colorectal cancer (PCCRC). It is known that the majority of PCCRCs arise from missed lesions (premalignant polyps or cancers]. or incomplete polypectomy.^7,8 Back-to-back colonoscopy studies show that 22% of all adenomas are missed,^9–14 and that there is a three- to sixfold variation in adenoma detection rates between endoscopists.^15,16 Even when polyps are found, removal may be incomplete: the Complete Adenoma REsection (CARE) study concluded that 10% of nonpedunculated polyps of 5–20 mm and 23% of nonpedunculated polyps of 15–20 mm were incompletely resected.¹⁷ Furthermore, low cecal intubation rates and poor bowel preparation regimens may explain the relative failure of colonoscopy to protect against proximal colorectal cancer that was found in many studies.^18–25 This results in clinically important differences in quality of care and patient outcomes: a recent study in the UK demonstrated a more than fourfold variation in PCCRC rates between hospitals.²⁶

In the upper GI tract, gastric cancers and precursor lesions are frequently missed: in one series, 7.2% of patients with gastric cancer did not have the lesion detected at endoscopy performed in the preceding 1 year. Of these cases, almost three quarters were felt to be due to endoscopist error.²⁷ Equally, in ERCP, which is one of the most complex and highest risk procedures performed regularly in endoscopy practice, there is evidence of wide variation in both completion and complication rates.^28–35

Performance measures

Providers and users of services can only know whether their service is delivering good quality care if it is measured. Performance measures are measurements that are used to assess the performance of a service or aspect of a service; other terms used for these include quality measures, quality indicators, key performance indicators, or clinical quality measures. Evidence-based performance measures provide endoscopists and endoscopy units, both often working in relative isolation, with a framework and benchmark against which they can assess their service.

Knowledge of the significant variation in quality between endoscopists does not improve quality per se, but setting minimum and target standards within these measures incentivizes improvement: when clinicians and services see their own performance data, they act to improve them. Open publication of performance measures also permit users of the service to assess quality for themselves, thus making better informed choices and further incentivizing improvements in healthcare. However, although open publication has potential benefits, it can cause unintended damage if handled poorly, for example if data are open to misinterpretation or inappropriate comparison. Thus it is important to consider both the benefits and risks of open publication for each case.

The provision of high quality endoscopic care is complex, involving myriad people, processes, and equipment. Healthcare professionals work hard to deliver this service, yet failure of any aspect may result in suboptimal care and poor health outcomes. Performance measures help a service to identify, appraise, and monitor the key steps in the process and the key outcomes, showing where systems are suboptimal and whether the service is providing high quality patient-centered healthcare.

Carefully constructed performance measures should allow providers to identify and address specific deficits in their service, resulting in better patient outcomes. Good performance measures should therefore correlate with an important health outcome. These measures should be evidence-based, clear, objective, reproducible, and realistic. They should also be practical to measure and meaningful for their target audience (for example endoscopists, patients, or healthcare providers). In an ideal construct, there should be a small number of carefully selected performance measures assessing all important aspects of the service (domains). Each measure assesses performance from a specific angle. Together they provide a holistic snapshot of the quality of the service. Some performance measures may relate to broad procedures (for example, cecal intubation rate), whereas others may relate to specific steps in a specific procedure (for example the optimal biopsy strategy for surveillance of Barrett’s esophagus).

Performance measures can be used to measure the quality of organizational structure, healthcare processes, or clinical outcomes. They can be applied in the pre-, intra- or post-procedural time periods.

• Structural measures reflect the conditions in which providers care for patients, in other words they reflect aspects of healthcare infrastructure. These measures can provide information about procedural volumes performed by a provider, staffing levels or, for example, whether a provider has adopted an electronic endoscopy reporting system.

• Process measures show whether actions proven to benefit patients are being completed. An example would be the percentage of patients requiring pre-procedure antibiotics who receive the correct antibiotic at the correct time.

• Outcomes measures analyze the actual results of care. These are generally the most important measures. An example would be the percentage of patients readmitted to hospital for a complication within 30 days of the endoscopic procedure.

Performance measures describe what to measure. However, it is usually desirable to take this further, identifying a minimum standard and a target standard within the measure. For example, it might be decided that cecal intubation rate is an important performance measure of colonoscopy; within this, a minimum standard might be set at 90% or 95%, with a target standard of 97%. Whereas performance measures will remain relatively static over time, the standards within such measures will be more dynamic, changing over time as techniques and technology improve. Moreover, the standards may vary according to procedure: for example, the minimum standard for adenoma detection rate will be higher for diagnostic colonoscopy performed because of fecal occult blood findings compared with colonoscopy prompted by symptoms. Occasionally no clear minimum standard currently exists for a performance measure (for example, patient comfort), yet its assessment may still be considered important. These are sometimes described as “auditable outcomes,” and it is hoped that in time, further research will help determine appropriate standards. Owing to small sample size, rates for rare events, such as missed cancers, may be best examined at endoscopy unit level rather than endoscopist level, whilst a qualitative review of each case is also performed (root cause analysis).

The terminology used in measuring quality can be confusing. A summary of terminology is presented in Table 1.

Table 1.

Terminology used in measuring quality

Term	Description/definition	Example
Domain	An area of clinical practice	Completeness of procedure, identification of pathology, management of pathology, complications, patient satisfaction
Performance measure	A measure that helps assess performance within a domain. Other terms used for this include quality measure, quality indicator, key performance indicator, or clinical quality measure. Can look at structure, process, or outcome.	Cecal intubation rate (CIR)
Minimum standard	A minimum defined level of performance within a performance measure	Minimum CIR standard is ≥90%
Target standard	A desirable/aspirational level of performance within a performance measure	Target CIR standard is ≥95%

The ESGE Quality Improvement Initiative

The ESGE Quality Improvement Committee (QIC) was instigated in 2013. Its aims are:

• To improve the global quality of endoscopy and the delivery of patient-centered endoscopy services

• To promote a unifying theme of quality of endoscopy within ESGE activities, achieved by collaborating with other ESGE committees and working groups and underpinned by a clear quality improvement framework

• To assist all endoscopy units and endoscopists in achieving these standards.

QIC committee membership comprises the QIC chairperson (M.R.), ESGE president and president-elect, chairs of the other three ESGE committees (guidelines, education and research) and chairs of QIC working groups.

A QIC strategy was developed to aid fulfilment of ESGE QIC aims. Quality improvement is a dynamic process and as such the strategy details will evolve over time, although the broad quality remit will not. An initial key objective was to help improve the quality of gastrointestinal endoscopy by producing a framework of performance measures for endoscopy, including quality of independent endoscopists and quality of endoscopy services (covering all aspects of the service including equipment, decontamination, waiting times, and patient experience), by developing robust, evidence-based performance measures. The aim of this was to set a minimum standard for individual endoscopists and for the endoscopy service, and to permit endoscopy units to measure their services against this patient-centered framework.

It was determined that such performance measures should be constructed using a rigorous evidence-based consensus process, incorporating a wide variety of stakeholders, including patients, from as wide a geographical area as possible. The aim was to delineate the core domains of a quality endoscopy service, to identify performance measures within each domain, and precisely to define and describe a small number of key performance measures covering each domain.

As the project fulfilled a key aim of the UEG Strategic Plan 2015–2018, ESGE approached UEG regarding potential collaboration and UEG agreed to this collaboration. Both ESGE and UEG co-funded the project and provided additional project governance.

The QIC committee created four working groups related to different areas of the gastrointestinal (GI) tract: upper GI, lower GI, pancreatobiliary, and small-bowel. A fifth “Endoscopy Service” working group was also created. An open call for expressions of interest (EOI) in participation was launched by ESGE, by emailing all individual members and all ESGE-affiliated endoscopy societies and by placing an article in the ESGE newsletter. A total of 90 EOIs were received from over 30 nations. The QIC committee nominated, approached, and appointed working group chairs and a meeting with these chairs was held to discuss the project in detail. Utilizing the list of EOIs, each working group chair established their working group membership, aiming to ensure as wide a geographical spread as possible, with between 10 and 20 members per GI tract group. Because of the nature of the Endoscopy Service group with regards to varying practice between nations, membership of this working group was deliberately larger and each ESGE-affiliated national endoscopy society was asked to nominate an individual to participate in the group, which comprised 34 members. No individual was permitted to be in more than one group. The American Society for Gastrointestinal Endoscopy (ASGE) was approached regarding collaborative involvement and agreed to provide input specifically into the small-bowel working group, along with overall comment or endorsement of the project output as appropriate.

The QIC committee contracted an expert team of methodologists to provide methodological support and to conduct the detailed literature searches (Literature Group). The Literature Group leader (C.S.) was co-opted onto the QIC committee for the duration of the project. To facilitate the program, a bespoke web-based platform was commissioned (ECD Solutions, USA). Within this platform, modules were created corresponding to the steps in the development process. All working group members had access to these modules, permitting both open and anonymized discussion around each aspect of the performance measure development. An expert in guideline methodology with significant prior experience of working with similar web-based platforms (C. Bennett) was commissioned to facilitate the integration of the information technology component.

Performance measures project process

A multistep process was developed by the QIC committee (Table 2). The Appraisal of Guidelines for Research and Evaluation II (AGREE II) tool was used to structure the guideline development process,³⁶ incorporating best practice from both the Scottish Intercollegiate Guidelines Network (SIGN) development processes and the National Quality Measures Clearinghouse (NQMC) of the United States of America. To ensure working group members had an understanding of guideline development methodology, all completed the SIGN online critical appraisal course (http://www.sign.ac.uk/methodology/tutorials.html; with permission).

Table 2.

Performance measures project: process steps

Establishment of QIC and project working groups

Declaration of conflicts of interest – all working group members

Complete SIGN online critical appraisal course – all working group members

Define the domains across all four GI fields (upper GI, small-bowel, pancreatobiliary, lower GI) and separately for Endoscopy Service (agreed by modified Delphi consensus process across all working groups)

Create PICOs, listing all key outcomes

Conduct literature search and construct evidence table

Create long-list of performance measures for each domain within each working group

Use ISFU checklist (Table 5) for each potential performance measure. Discard inferior performance measures, and where no performance measure exists within a domain, construct appropriate performance measure by modified Delphi consensus process

Determine final performance measures – modified Delphi consensus process

Develop descriptive framework for each performance measure (Table 6). Review, tabulate and GRADE evidence for minimum/target standards within each performance measure

Review and harmonization of performance measures across all five working groups

Highlight areas for future research based on gaps in evidence identified during this process

Identify training/education needs

Review by ESGE, UEG, national societies, and patient groups for comment and consensus

Final amendments – modified Delphi process including ESGE QIC committee

QIC, Quality Improvement Committee; SIGN, Scottish Intercollegiate Guidelines Network; GI, gastrointestinal; PICOS, population/patient, intervention, comparison, outcome, study design; ISFU, Importance, Scientific acceptability, Feasibility, and Usability; GRADE, Grading of Recommendations Assessment, Development and Evaluation; ESGE, European Society of Gastrointestinal Endoscopy: UEG, United European Gastroenterology.

A preliminary meeting for all working group members was held at the UEG Week conference in Vienna, October 2014. The project was explained in detail and each working group proposed potential domains for endoscopy. After open discussion, a draft single set of domains, unified across all the four GI tract areas, was constructed and voted on using a modified Delphi consensus process, as described in Table 3,³⁸ If consensus was not reached initially, further discussion and voting was performed to re-evaluate and modify proposed domains until consensus was reached. The agreed domains for the GI tract working groups included completeness of procedure, identification of pathology, management of pathology, complications, procedure numbers, and patient experience.

Table 3.

Modified Delphi consensus process

Consensus voting was conducted through the website. Consensus was reached using a modified Delphi technique. Each working group member anonymously scored their level of agreement with draft measures using a 1 to 5 scale: 1 = Strongly agree, 2 = Agree, 3 = Neither agree nor disagree, 4 = Disagree, 5 = Strongly disagree.

Space was provided to include comments and additional references that were felt to require consideration. Commenting was mandatory for undecided or disagree votes.

At least 80% agreement (scores of 1 or 2) was required for consensus to be reached. Where consensus was not reached, measures were reviewed in light of comments made and any additional evidence identified, and were adjusted if required. Further voting rounds then took place for these measures.

If 80% agreement was not reached after a maximum of three rounds of voting, consensus was considered reached if >50% of participants voted in favor and <20% voted against the measure, in accordance with the GRADE process.37 Failure to meet this criterion resulted in the measure being discarded.

Each working group developed an exhaustive list of potential areas for literature review, using the PICOS (Population/Patient, Intervention, Comparison, Outcome, Study design) process.^39–41 The questions were focused on the assessment of the relationship between specific indicators and procedure outcomes (e.g. completion rate) or patient outcomes (e.g. interval cancer rate, change in clinical management). PICOS were reviewed by the Literature Group and revisions made until a final precisely defined list was reached. The PICOS components of each prioritized question were used by the Literature Group to define specific keywords for the comprehensive bibliographic searches. If more than one comparison was deemed to be relevant, the results of each comparison were reported.

Searches were performed on the Cochrane Central Register of Controlled Trials (CENTRAL), Medline and Embase, from 1 January 2000 to 28 February 2015, using MESH terms and free-text words, without language restriction. In the first instance systematic reviews were searched. If updated systematic reviews addressing the PICOS questions were retrieved, the search for primary studies was limited to those studies published after the last search date of the most recently published systematic review. If no systematic reviews were found, a search of primary studies since 2000 was performed. In order to avoid repetition or double counting of primary studies, where a literature search retrieved many systematic reviews addressing the same PICOS question, only the best systematic review, based on the evaluation of their methodological quality, update of the bibliographic search, level of overlapping, and quality of evidence of included primary studies, was considered for data extraction.

A hierarchy of the study designs to be considered for each type of question (e.g. on effectiveness, diagnostic accuracy, acceptability, and compliance) was produced by the epidemiologists of the Literature Group. For effectiveness questions, randomized controlled trials were considered as the best source of evidence and were searched in the first instance. For diagnostic accuracy questions, cross-sectional studies with verification by reference standard were considered as the best source of evidence.

The risk of bias of included studies was assessed using the following validated checklists:

• systematic review: AMSTAR (Assessing the Methodological Quality of Systematic Reviews) checklist⁴²

• randomized controlled trials: The Cochrane Collaboration’s tool for assessing risk of bias in randomized trials⁴³

• cohort studies, case–control studies and cross-sectional surveys: Newcastle-Ottawa Scale⁴⁴

• diagnostic accuracy studies: QUADAS 2 (Quality Assessment Tool for Diagnostic Accuracy Studies 2) checklist⁴⁵

• interrupted time series analysis: criteria suggested by the Cochrane Effective Practice and Organisation of Care Review Group.⁴⁶

The draft results of the bibliographic search and of the selection process produced by the Literature Group were reviewed by the clinical experts of the working groups, to determine whether the inclusion of additional evidence or the exclusion of nonrelevant papers was required. Once necessary revisions were made, for each question or group of questions pertaining to the same topic, the Literature Group provided an evidence table with the main characteristics of each included study (study design, objective of the study, comparisons, participant characteristics, outcome measures, results, risk of bias). They also provided a summary document with a description of the search strategy used for each database, the overall number of titles retrieved, and the number of potentially relevant studies acquired in full text; the number of studies finally included was given, as well as a synthesis of their characteristics and risk of bias, and of their results, overall conclusions, and quality of evidence.

The Grading of Recommendations Assessment, Development and Evaluation (GRADE) tool was used to evaluate both the quality of evidence and the strength of recommendations made (Table 4).^48,49 The GRADE system specifically separates the quality of evidence from the strength of a recommendation: whilst the strength of recommendation may often reflect the evidence base, the GRADE system allows for occasions where this is not the case, for example where there appears to be good reason to make a recommendation in spite of an absence of high quality scientific evidence such as a large randomized controlled trial.

Table 4.

An overview of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system.⁴⁷

GRADE: Strength of evidence
High quality:
Further research is very unlikely to change our confidence in the estimate of effect
Moderate quality:
Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate
Low quality:
Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate
Very low quality:
Any estimate of effect is very uncertain
GRADE: Strength of recommendation
Recommendations can be categorized as either Strong or Weak. Recommendations involve a trade-off between benefits and harms. Those making a recommendation should consider four main factors:
• The trade-offs, taking into account the estimated size of the effect for the main outcomes, the confidence limits around those estimates, and the relative value placed on each outcome
• The quality of the evidence
• Translation of the evidence into practice in a specific setting, taking into consideration important factors that could be expected to modify the size of the expected effects, such as proximity to a hospital or availability of necessary expertise
• Uncertainty about baseline risk for the population of interest. If there is uncertainty about translating the evidence into practice in a specific setting, or uncertainty about baseline risk, this may lower our confidence in a recommendation.

Once the literature review was completed, initial draft evidence statements with comprehensive supporting documentation were uploaded onto a customized web platform, for all working group members to review and comment in a modified Delphi process (see Table 3), to allow modification and to identify additional references. Where necessary, further literature reviews were undertaken and further revisions made in subsequent voting rounds.

From the final evidence construct, the working group chairs identified draft performance measures, aiming for a small number of key measures per domain. Where no measure had been identified within a domain, the working group was permitted to construct one by consensus if deemed clinically appropriate. Once the key performance measures had been identified, each measure was evaluated using the ISFU (Importance, Scientific acceptability, Feasibility, and Usability) framework described by the National Quality Measures Clearinghouse (Table 5).⁵⁰ Measures which did not meet the criteria were discarded. The modified Delphi process was then used to reach consensus on these performance measures.

Table 5.

Importance, Scientific acceptability, Feasibility, and Usability (ISFU) system, customized and adapted to our working group needs

Importance to measure and report	Extent to which the specific measure focus is evidence-based, important to making significant gains in healthcare quality, and improving health outcomes for a specific high priority aspect of healthcare where there is variation in or overall less-than-optimal performance. Measures must be judged to meet all subcriteria to pass this criterion and be evaluated against the remaining criteria.	1a. Evidence base The measure focus is evidence-based: • Health outcome: a rationale supports the relationship of the health outcome to processes or structures of care. • A systematic assessment and grading of the quantity, quality, and consistency of the evidence that the measured structure, process or intermediate clinical outcome leads to a desired health outcome. 1b. Performance gap Demonstration of quality problems and opportunity for improvement 1c. High priority A high priority aspect of healthcare.
Scientific acceptability of measure properties	Extent to which the measure, as specified, produces consistent (reliable) and credible (valid) results about the quality of care when implemented. Measures must be judged to meet the subcriteria for both reliability and validity to pass this criterion and be evaluated against the remaining criteria.	2a. Reliability The measure is well defined and precisely specified so it can be implemented consistently and allows for comparability. 2b. Validity The measure specifications are consistent with the evidence. Target population and exclusions are supported by the evidence. Validity testing demonstrates that the measure correctly reflects the quality of care provided, adequately identifying differences in quality. Where an evidence-based risk-adjustment strategy is specified, it has demonstrated adequate discrimination and calibration. Analysis of computed measure scores demonstrates that scoring allows for identification of statistically significant and practically/clinically meaningful differences in performance. If multiple data sources/methods are specified, there is demonstration they produce comparable results. For measures susceptible to missing data, analyses identify the extent and distribution of missing data (or nonresponse) and demonstrate that results are not biased due to it and how the specified handling of missing data minimizes bias. 2c. Disparities If disparities in care have been identified, measure specifications, scoring, and analysis allow for identification of disparities through stratification of results.
Feasibility	Extent to which the specifications, including measure logic, required data that are readily available or could be captured without undue burden and can be implemented for performance measurement.	3a. For clinical measures, the required data elements are routinely generated and used 3b. The required data elements are available in electronic sources, or a credible path to electronic collection is specified. 3c. Demonstration that the data collection strategy can be implemented
Usability and use	Extent to which potential audiences (e.g., consumers, purchasers, providers, policymakers) are using or could use performance results for both accountability and performance improvement to achieve the goal of high quality, efficient healthcare for individuals or populations.	A credible rationale describes how the performance results could be used to further the goal of high quality, efficient healthcare for individuals or populations.
Comparison to related or competing measures	If a measure meets the above criteria and there are endorsed or new related measures (either the same measure focus or the same target population) or competing measures (both the same measure focus and the same target population), the measures are compared to address harmonization and/or selection of the best measure.	Consider multiple measures in a domain if: The measure is harmonized with related measures or multiple measures are justified. Consider replacing existing measure if: The measure is superior to existing measures

A detailed descriptive framework was then constructed for each measure meeting the ISFU criteria, as described in Table 6.⁵¹ Quality standards (minimum and target) were identified within each performance measure. Additional literature searches were performed where necessary. Where no evidence-based standard was identified, the working group was permitted either to agree on a suitable standard by consensus, or to state “no current standard defined.”

Table 6.

Customized and adapted descriptive framework for each final performance measure

Performance measure	[name]
Description	Provide a concise summary statement of performance measure
Domain	[domain name]
Category	Structure/Process/Outcome
Rationale	Explain the importance of the measure
Evidence for performance measure	Use GRADE system for evidence base and for strength of recommendation
Details	Clearly describe: Target population (denominator) Identification of those from the target population who achieved the specific measure focus (numerator, target condition, event, outcome) Measurement time window Exclusions Risk adjustment/stratification Definitions Data source and feasibility Consider handling of missing data Specifications for composite performance measures include: component measure specifications (unless individually endorsed); aggregation and weighting rules; handling of missing data; standardizing scales across component measures; required sample sizes
Scoring	Describe how the performance measure is calculated (e.g. mean/median, count, ratio, rate/proportion) Indicate if stratification/case mix adjustment or weighting required Frequency of calculation. Describe level of analysis (e.g. individual endoscopist – cecal intubation rate; or service level – bowel preparation quality)
Minimum/target standards	Describe minimum/target standards State “no current standard defined” where none exists Describe how score should be interpreted relative to the minimum/target standard Describe whether the standard includes any tolerance for any factors Describe action that should be taken when performance does not reach minimum standard

Along with the final list of precisely defined key performance measures, the working groups compiled a longer list of other performance measures that had been identified during the development process, a list of areas with weak evidence base for priority research, and a list of training/educational needs. The final draft was then reviewed by the ESGE QIC Committee and the ESGE Governing Board. Finally, review and approval was obtained from ESGE-affiliated national societies, UEG, ASGE, and patient groups.

The ESGE quality improvement vision

ESGE and UEG have a vision to create a thriving community of endoscopy services across Europe, collaborating with each other to provide high quality, safe, accurate, patient-centered, and accessible endoscopic care. Whilst the boundaries of what can be achieved in advanced endoscopy are continually expanding, we believe that one of the most fundamental steps to achieving our goal is to raise the quality of everyday endoscopy. The development of robust, consensus- and evidence-based key performance measures is the first step in this vision.

Implementing performance measures, along with additional measures such as structured training programs, can result in significant improvement in endoscopy quality. In the UK for example, a decade of quality improvement initiatives resulted in cecal intubation rate improving from 76.9% to 92.3%.⁵²

Having a performance measure does not result in improved health outcomes per se: in order to improve quality, it is essential to measure local performance regularly against this benchmark. Services and individuals are unlikely to improve unless they are aware of their performance and how it compares with benchmark performance measures. Measuring allows the identification of potential underperformance, which provides an opportunity for discussion and support for the endoscopist. In addition, the simple act of monitoring a service will improve performance (the “Hawthorne effect”): it is powerful, essentially free, and results in improved quality of patient care.

The standardization of performance measure definitions and measurement methodology is crucial to permit comparative assessment. Quality improvement requires political will. At a local level, it requires support from hospital management. Whilst not essential, the best examples of quality improvement in endoscopy have also had commitment from, indeed have often been led by, regional or national authorities and we call upon such organizations to share responsibility for and to facilitate this program. The implementation of appropriate information technology infrastructure, based around electronic endoscopy reporting systems, is an important step in allowing timely data collection and automated, standardized performance measure reporting.

A strong case can be made for setting a minimum number of procedures per endoscopist per year. Firstly, a large sample size increases the accuracy of the performance measurement (i.e., it reduces the probability that apparent underperformance is a chance event). Secondly, there is evidence that endoscopy proficiency increases with increasing number of procedures performed, and that endoscopy complications are more common with endoscopists who perform fewer procedures per year¹; this is also well described in many other clinical areas such as surgery.⁵³ A trend towards fewer endoscopists each performing more procedures may be appropriate, and setting a minimum number of procedures per year for endoscopists may be one strategy to improve quality.

It is important that we help endoscopists with lower levels of performance to improve. Quality assurance should be about improvement, not punishment. One of the biggest gains in endoscopy quality improvement would be to raise the standards of the lower performers to above minimum quality standard thresholds. Various organizations have developed structured processes for the management of underperforming endoscopists, and experience shows that when handled sensitively but robustly, most endoscopists embrace such support. However, there may at times be barriers to the uptake of endoscopy quality improvement by individuals and even services, ranging from complacency (“I’m fine and don’t need to measure”) to fear that one’s abilities might be demonstrated to be suboptimal. The latter may be particularly relevant if there are financial or service imperatives to continue with the status quo. Nevertheless, we owe it to our patients to overcome these barriers to ensure that endoscopy is of the highest quality.

ESGE and UEG have identified quality of endoscopy as a major priority. We recommend that all units develop mechanisms for audit and feedback of endoscopist and service performance, using the ESGE performance measures that will be published in future issues of Endoscopy over the next year. Regional and national organizations have a responsibility to support and, where required, provide resources for such quality improvement initiatives. We urge all endoscopists and endoscopy services to prioritize quality and to ensure that these performance measures are implemented and monitored at a local level, so that we can provide the highest possible care for our patients.

Competing interests

Competing interests: M. Rutter’s department receives research funding from Olympus for a colitis surveillance trial (2014 to present). C. Senore’s department receives PillCam Colon devices from Covidien-Given for study conduct, and loaner Fuse systems from EndoChoice. R. Bisschops has received: speaker’s fees from Covidien (2009–2014) and Fujifilm (2013); speaker’s fee and hands-on training sponsorship from Olympus Europe (2013–2014); speaker’s fee and research support from Pentax Europe; and an editorial fee from Thieme Verlag as co-editor of Endoscopy.

R. Valori is a director of Quality Solutions for Healthcare, a company providing consultancy for improving quality and training in healthcare. C. Spada has received training support from Given Imaging (2013 and 2014). M. Bretthauer receives funds from Thieme Verlag for editorial work for Endoscopy. C. Bennett owns and works for Systematic Research Ltd, and received a consultancy fee from ESGE to provide scientific, technical, and methodological expertise for the present project. C. Hassan has received equipment on loan from Fujinon, Olympus, Endochoice, and Medtronic; and consultancy fees from Medtronic, Alpha-Wasserman, Norgine, and EndoChoice. C. Rees’s department receives research funding from Olympus Medical, ARC Medical, Aquilant Endoscopy, Almirall, and Cook (from 2010 to the present). M. Dinis-Ribeiro receives funds from Thieme Verlag for editorial work for Endoscopy; his department has received support from Olympus for teaching protocol (from August 2014 to July 2015). T. Ponchon has received: advisory board member’s fees from Olympus, Ipsen Pharma, and Boston Scientific (2014 and 2015) and from Cook Medical (2014); speaker’s fees from Fujifilm, Ipsen Pharma, and Olympus (2014 and 2015) and from Covidien (2014); training support from Ferring (2014); and research support from Boston Scientific and Olympus (2014 and 2015). P. Fockens has been receiving consulting support from Olympus, Fujifilm, Covidien, and Creo Medical. L. Aabakken, C. Bellisario, D. Domagk, T. Hucl, M. Kaminski and S. Minozzi, have no competing interests.