High quality performance is important to prevent colorectal cancers (CRCs) after colonoscopy. Therefore, we endoscopists are measured every day on our rates of cecal intubation, detection of polyps and adenomas, and patient satisfaction. That is great, and we support it. It helps us to understand our own performance in comparison to our peers, increases patient benefits, and reduces harms. After all, our endoscopies are there for the benefit of the people we treat: our patients.

Occasionally, a patient who we have recently scoped gets cancer. This is not good, but it is inevitable because first the risk of cancer after colonoscopy can never be zero and second sometimes the cancer is not causally related to the performance of the colonoscopy (although good quality reduces the risk). Thanks to the modern endoscopy quality improvement systems we have set up, many endoscopy services are now able to track a CRC developing after a colonoscopy back to the colonoscopy that one of us performed some time prior to the diagnosis. The question is: how do we learn from such unwanted events to prevent them in the future? The current issue of Endoscopy features a study that sheds light on this important question by disentangling the reasons for CRCs after colonoscopy [1].

The study empaneled data from the nationwide CRC screening program in the Netherlands and constructed a cohort of patients who were diagnosed with CRC within about 5 years of their colonoscopy for a positive FIT screening test between 2014 and 2016 [1]. Out of over 116000 patients who were scoped after a positive FIT, 432 patients were diagnosed with CRC a median of 31 months after the colonoscopy (CRC diagnoses in the first 6 months after the colonoscopy were not included – this is common and well reasoned in studies of CRC after colonoscopy).

The researchers followed guidance from the World Endoscopy Organization (WEO) and categorized the cancers into four types [2]:

cancers diagnosed before the date of the recommended surveillance colonoscopy

cancers diagnosed at the recommended surveillance colonoscopy

cancers diagnosed after the recommended surveillance colonoscopy

cancers diagnosed after the recommended surveillance colonoscopy in patients where surveillance was not performed owing to co-morbidity.

The WEO guidance document terms the first category “interval post-colonoscopy CRC” (PCCRC), and the three others “non-interval PCCRC.” Most cancers in the Dutch study (52%) were in the first category.

The most important part of the Dutch study is the thoroughly performed root-cause analysis of each cancer. It indicates that 74% of all cancers diagnosed before the date of the recommended surveillance colonoscopy are likely due to overlooked lesions. Incomplete removal or polyps not being removed were common reasons for the other cancer types, but the numbers were small and there is uncertainty around the observed point estimates. The findings of this new study are consistent with other recent evidence that many interval cancers are avoidable. A recent UK cohort found that approximately 70% of cancers after colonoscopy may be due to missed or incompletely removed polyps or lesions [3].

A couple of features in the new Dutch study deserve further consideration. First, the authors report a lower risk of cancer after colonoscopy compared with FIT-screening studies in some other European countries. Whether this is because of different background cancer risks in the populations, different FIT cutoffs in the national screening programs, or better training of endoscopists in the Netherlands is difficult to disentangle. Whether the Dutch results can be translated to colonoscopy screening programs instead of FIT programs is uncertain, but the absolute risks are likely to be smaller owing to the different risk profiles of individuals who undergo colonoscopy in colonoscopy screening programs compared with those within FIT programs.

Second, the authors point out that the WEO guidance document requires an update to include possible missed advanced serrated polyps as a reason for interval cancer after colonoscopy. Currently, only for adenomas in the same segment are inferences allowed to be made about the reason for interval cancer in root-cause analyses [2].

Third, caution is warranted when comparing cancer stage for the different types of interval cancer. Because cancers detected at surveillance colonoscopy are detected earlier than clinically symptomatic cancers, such analyses are potentially invalid owing to lead-time bias and likely also overdiagnosis bias. Finally, if one wants to investigate the prognosis of interval cancer after screening, the most valid comparison group is clinically detected cancers, and not screening-detected cancer [4], with the latter comparison biased by lead-time and overdiagnosis.

Structured quality work in endoscopy started in the early 2000s after there was a realization of poor endoscopy services in the National Health Service (NHS) in England [5]. Most of the World has since followed the guidance of the English colonoscopy quality visionaries who implemented modern quality improvement systems in colonoscopy in the NHS. The focus in the first decade has been to carve out the best indicators for good quality (so-called key performance indicators) and to find out how to train endoscopists properly to achieve good quality [6]. This has been instrumental in professionalizing the field and ensuring good service for CRC screening, where the margins of benefits are small and avoidable harms due to poor quality are not acceptable.

More recently, root-cause analyses have become an important part of quality assurance programs in colonoscopy. The Dutch study is an excellent example of their importance in understanding causal relationships for rare adverse events after colonoscopy, such as CRC. Root-cause analyses involve dedicated interdisciplinary teams who perform structured reviews of individual adverse events using chart reviews, interviews with involved personnel, and leadership engagement [7]. Importantly, root-cause analyses are open to any findings of causation (they are agnostic by design), they emphasize system-level responsibility, and they include clear proposals of change to improve performance and reduce the likelihood of repeat adverse events [7].

Finally, a kind request to our endoscopy community: if we want to make ourselves understood, we must use easy-to-understand language and abbreviations. Our current terminology for cancer after colonoscopy needs improvement. First, the abbreviation PCCRC should be abandoned. A good way to prevent easy understanding and buy-in for good ideas and concepts is to create five-letter abbreviations (too many letters), with many similar letters after each other (in this case Cs), and no hyphen or other hint to ease understanding for struggling readers. Second, we should re-think whether we need all the subcategories for cancers after colonoscopy [1] [2]. There are methodical problems, and they may not be all that interesting to differentiate for the learning from them [4]. I suggest keeping it simple and in line with what other fields in medicine are using, for instance the term “interval cancer” is well established in other fields in medicine. It is simple, short, and requires no abbreviation. Yes, it does require a by-line for further explanation and definition, but that is easily accommodated and it does not inhibit immediate understanding. After all, easy understanding is key to improvement.

Article published online:
13 March 2024

© 2024. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany