Quality assurance in colonoscopy: is case mix a problem?

The importance of colonoscopy quality assurance measures was properly embraced when the European Union in 2003 recommended organized colorectal cancer (CRC) screening programs in member states provided quality assurance monitoring at all levels of the service [1]. Independently, or in the wake of this, the first two colonoscopy quality assurance programs were launched in the United Kingdom and Norway in the autumn of 2003 [2] [3]. Many countries followed suit and convincing results, mostly from CRC screening studies, emerged. After some years, reduced interval cancer rates were found to be correlated with adenoma detection rates (ADR) and polypectomy [4] [5].

“The study by Nass et al. is a call for awareness of the risk for performance misclassification if possible case-mix differences are not considered.”

As adequate ADR is both a good measure of colonoscopy quality, at least in a screening setting, and has a proven effect in cancer risk reduction, many consider ADR to be the most important of the seven quality indicators defined by the European Society of Gastrointestinal Endoscopy [6]. However, ADR is not always readily available because electronic pathology reporting is seldom integrated with endoscopy reports in hospital electronic medical records. Cecal intubation rate (CIR), adequate bowel preparation rate (ABPR), detection rate for any polyp (PDR) or polyps ≥ 5 mm in diameter (PDR-5 mm), withdrawal time, and patient-reported pain are some of the more surrogate indicators that are more easily accessible; these measures often have value in themselves, such as CIR and patient-reported experience and outcome measures.

There are concerns that case-mix differences may lead to erroneous categorization of performance of endoscopy centers and individual endoscopists. In this issue of Endoscopy, Nass et al. report on case-mix adjustments in a national registry study from the Netherlands, including 51 endoscopy centers and more than 363 000 colonoscopies (approximately 183 000 diagnostic, 100 000 surveillance, and 80 000 work-up colonoscopies after a positive fecal immunochemical test [FIT]) [7]. Patient age and sex, American Society of Anesthesiology (ASA) classification, and indication for colonoscopy were significant case-mix factors for CIR and ABPR, and they varied between centers. For indication work-up after FIT screening, age, sex, ASA classification, and center were again significant factors for CIR. This study is a call for awareness of the risk for performance misclassification if possible case-mix differences are not considered; case-mix differences may not present a problem, but they should be considered.

Performance data per center may reveal substandard performance that requires improvement. The actions necessary to address performance issues are the responsibility of the head of unit or center. The publication from Nass et al. is based on data from the nationwide Dutch Gastrointestinal Endoscopy Audit database, which now also allows reporting on some quality assurance data per endoscopist by using a unique endoscopist identification number. If quality assurance registers cannot provide performance data per endoscopist, this may be a lost opportunity for efficient up-skill initiatives. Should all endoscopists in the unit be sent to up-skill courses? Possibly, if we do not have access to relevant data at the endoscopist level; however, such action may have a negative impact on waiting times for colonoscopies.

The variables subject to adjustment in analyses, and targeted improvements if possible, are patient demographics (e. g. sex, age for data analysis), organization (e. g. bowel preparation strategies), and endoscopist performance (skills improvement), as well as team functioning, which has barely been addressed in the literature. Reporting relevant quality assurance data at the individual patient level but not at the individual endoscopist level appears half-hearted when we know that performance at the center level depends on the combined performance of the endoscopists.

Variations in ADR, CIR, PDR, PDR-5 mm, and ABPR between centers may be due to differences in patient demographics, comorbidity, and bowel preparation, but some of the variations may equally be due to differences in endoscopists’ skills within each endoscopy unit. Furthermore, nurses within the endoscopy team, though not performing the colonoscopy itself, have been shown to influence quality, emphasizing the important role of team functioning [8].

In a study of more than 100 000 colonoscopies (including about 20 000 screening-related colonoscopies) performed by 102 endoscopists, the authors concluded that adjusting for all available case-mix variables only had a minimal effect on the ADR ratings of individual endoscopists [9]. The importance of case-mix adjustment may depend on how the endoscopy services are organized, and on the recruitment mechanisms for doctors and staff into private or public services and into diagnostic, screening or mixed diagnostic/screening services, respectively. In a study within an exclusively public colonoscopy service performing diagnostic and screening colonoscopies, an observed difference in PDR-5 mm could be explained by the endoscopist, whereas patient age, sex, CIR, bowel cleansing or indication for colonoscopy could not explain the observed difference [10]. Therefore, in this particular setting, it would be relevant for some endoscopists to consider measures to improve their ability to detect polyps rather than assuming that substandard performance is due to case-mix differences.

The present study by Nass et al. is a warning that quality assurance measures may require adjustment for case mix when comparing between centers. For service improvements, however, quality assurance registers should provide data per endoscopist as well as per endoscopy center, in order to identify individual endoscopists who need to improve their performance. If endoscopist exposure poses a problem, this may be overcome by coded identity known only to the individual endoscopist, thus leaving it to the endoscopist to decide whether to share this information with peers and colleagues in order to improve center performance.

Publication History

Article published online:
14 December 2021

• References

• 1 Boyle P, Autier P, Bartelink H. et al. European Code Against Cancer and scientific justification: third version (2003). Ann Oncol 2003; 14: 973-1005
• 2 Hoff G, Bretthauer M, Huppertz-Hauss G. et al. The Norwegian Gastronet project: continuous quality improvement of colonoscopy in 14 Norwegian centres. Scand J Gastroenterol 2006; 41: 481-487
• 3 Joint Advisory Group on GI Endoscopy. Available at (Accessed 23.10.2021): www.thejag.org.uk
• 4 Kaminski MF, Regula J, Kraszewska E. et al. Quality indicators for colonoscopy and the risk of interval cancer. N Engl J Med 2010; 362: 1795-1803
• 5 Corley DA, Jensen CD, Marks AR. et al. Variation of adenoma prevalence by age, sex, race, and colon location in a large population: implications for screening and quality programs. Clin Gastroenterol Hepatol 2013; 11: 172-180
• 6 Kaminski MF, Thomas-Gibson S, Bugajski M. et al. Performance measures for lower gastrointestinal endoscopy: a European Society of Gastrointestinal Endoscopy (ESGE) Quality Improvement Initiative. Endoscopy 2017; 49: 378-397
• 7 Nass KJ, van der Vlugt M, Elfrink AKE. Case-mix adjustment to compare colonoscopy performance between endoscopy centers: a nationwide registry study. Endoscopy 2022; 54: 455-462
• 8 Holme Ø, Pedersen IB, Medhus AW. et al. Endoscopy assistants influence the quality of colonoscopy. Endoscopy 2018; 50: 871-877
• 9 Jensen CD, Doubeni CA, Quinn VP. et al. Adjusting for patient demographics has minimal effects on rates of adenoma detection in a large, community-based setting. Clin Gastroenterol Hepatol 2015; 13: 739-746
• 10 Hoff G, Botteri E, Høie O. et al. Polyp detection rates as quality indicator in clinical versus screening colonoscopy. Endosc Int Open 2019; 7: E195-E202