The German colonoscopy screening program: the quality of a quality register

Colorectal cancer (CRC) is a worldwide health issue with 1.2 million new cases annually and a 5-year survival rate of approximately 50 %. Symptoms are unspecific and often appear late in its clinical course; however, only early diagnosis with surgery offers the prospect of being cured. Therefore, the European Union has recommended CRC screening programs for its member states since 2003, provided that adequate quality assurance programs are also implemented. This is in line with leading authorities on quality assurance in general, which state that “policy makers in research should reliably set aside a budget for research on quality and safety of care and implementation of change whenever they examine a new, relevant, clinical care topic” [1].

Regardless of the screening modality used, the quality of colonoscopy is the cornerstone of all CRC screening. Colonoscopy may be used as the primary screening modality, such as in the German and Polish colonoscopy screening programs, or for work-up following a positive primary flexible sigmoidoscopy, such as in the UK and Italy, or following a positive fecal occult blood test, which is used in most European countries.

Screening is a balance between benefits (reduced CRC incidence and mortality when compared with treatment of symptomatic disease) and harms (complications, perceived burden, psychosocial events). Keeping a good track of colonoscopy complications is a major task for quality assurance programs, particularly as in many countries the public demands that quality assurance data are freely available.

In the current issue of Endoscopy, Rösch et al. have addressed the quality of registration data in the German colonoscopy screening program, which has been running since 2002 and has accumulated data from 4 million screening colonoscopies. The study was a time-limited prospective study, which involved comparing separate parallel study registration with the standard registration of the screening program. This was done during an 18-month period in 19 Berlin screening centers and produced a consecutive series of 12856 screening colonoscopies, of which 12 150 were reported to the standard registry and 12 134 to the study database. Key parameters recorded were colonoscopy completion rate, adenoma detection rate, and complications. Study case report forms (CRFs), which were completed separately by endoscopists and patients, were used for documentation of data for the study database. The endoscopists were not aware of the intention to compare CRF-based study data with routine registry data until after completion of the study.

Registry and study data were similar with respect to colonoscopy completion and adenoma detection rates, but complications were reported more than three times as often in the study compared with the registry. Registry under-reporting was maintained even when patient feedback was not included and when complications were restricted to only those that required hospital admission. Data on complications entered into the standard registry and the study database were primarily provided by the same endoscopists, suggesting that a stronger focus on complications in the study by using two CRFs to capture early and late events, respectively, may have improved data quality. In the standard registry, there were no timelines for registration of complications, no patient feedback, and audits were limited to annual feedback requests for cancer details and to complications registered as “requiring hospital admission.”

This is a very important and timely study of the quality of quality assurance data. The main purpose of medical registries containing quality assurance data is to provide information that will guide efforts to improve the quality of the service provided. A prerequisite for the success of a quality assurance registry is trust in the quality of the data [2]; no one will invest resources into improving standards if the data being used cannot be trusted. At the same time, there is a general lack of funding for research on quality assurance [1]. Public demands for freely available quality data may improve this situation.

The authors concluded that, although there is a need for better definition of complications and timelines for reporting, endoscopist and patient-based self-reporting of complications may not be the best way to report complications. Registry linkage to hospital records and other databases may provide more reliable data but may be hampered by legal restrictions in many countries that limit the possibilities for data linkage to secure high-quality quality assurance registries. Many registries are dependent on patient consent and voluntary registration by the healthcare provider, which leaves such registries very prone to selection bias and we may end up with “feel-good” registries that may fail to disclose a need for quality improvement [3] [4] [5], and therefore benefit no one, least of all patients. Therefore, there is a strong ethical case for exempting informed consent from research on medical records and for facilitating linkage for quality assurance purposes [6].

The authors also point out that “a mandatory inclusion of such [patient] feedback information into a nationwide quality assurance program is probably unrealistic.” Nevertheless, this has been achieved in some countries. Patient feedback has been used as a standard and important part of the Norwegian Gastronet colonoscopy quality assurance registry since it was launched in 2003 [7]. However, it is resource demanding when based on paper records. Within the limited resources granted for Gastronet, conversion to web-based solutions is mandatory for continued maintenance of this important part of patient participation in quality work. The authors’ conclusion on this point is therefore fully understandable.

In conclusion, this is a milestone study on the work of quality assurance in gastrointestinal endoscopy. It should be a prompt for many institutions to invest effort and resources in their registries so that data may be reliably used for quality purposes.

• References

• 1 Grol R, Berwick DM, Wensing M. On the trail of quality and safety in health care. BMJ 2008; 336: 74-76
• 2 van der Veer SN, de Keizer NF, Ravelli AC et al. Improving quality of care. A systematic review on how medical registries provide information feedback to health care providers. Int J Med Inform 2010; 79: 305-323
• 3 Seip B, Bretthauer M, Dahler S et al. Sustaining the vitality of colonoscopy quality improvement programmes over time. Experience from the Norwegian Gastronet programme. Scand J Gastroenterol 2010; 45: 362-369
• 4 Tu JV, Willison DJ, Silver FL et al. Impracticability of informed consent in the Registry of the Canadian Stroke Network. N Engl J Med 2004; 350: 1414-1421
• 5 Hoff G. Quality assurance in gastroenterology: QA in research, and research in QA. Best Pract Res Clin Gastroenterol 2011; 25: 427-434
• 6 Miller FG. Research on medical records without informed consent. J Law Med Ethics 2008; 36: 560-566
• 7 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