Endoscopy 2018; 50(02): 93-95
DOI: 10.1055/s-0043-124871
© Georg Thieme Verlag KG Stuttgart · New York

Why don’t endoscopists follow guideline recommendations and how can we improve adherence?

Referring to Gessl I et al. p. 119–127
Philip Schoenfeld
Gastroenterology Section, John D. Dingell VA Medical Center, Detroit, Michigan, United States
› Author Affiliations
Further Information

Publication History

Publication Date:
29 January 2018 (online)

Current colorectal cancer (CRC) guidelines from the European Society of Gastrointestinal Endoscopy (ESGE) recommend a 10-year interval in low-risk patients with 1 – 2 small (< 10 mm) adenomas, and a 3-year interval for intermediate-risk patients with 3 – 4 small adenomas or at least one large (10 – 20 mm) adenoma, adenoma with high grade dysplasia or tubulo-villous adenoma [1]. If surveillance colonoscopy is repeated at shorter intervals, patients are subjected to the risks and inconveniences of colonoscopy more frequently, and the procedure is not cost-effective. Unfortunately, endoscopists in the United States frequently recommend surveillance colonoscopy sooner than recommended by guidelines [2] [3] [4]. In this issue of Endoscopy, Gessl et al. [5] report a similar practice by Austrian endoscopists.

“... it’s crucial to educate endoscopists that a 10-year interval is not too long in low-risk patients as long as a high quality colonoscopy is performed.”

Gessl et al. used the Austrian Certificate of Quality for Screening Colonoscopy (CQSC) database [6] to assess adherence to guideline recommendations for surveillance colonoscopy during 2011 – 2014. Adherence was less than 10 % in the low-risk group and less than 36 % in the intermediate-risk group. Most patients were instructed to return sooner than recommended by guidelines. Furthermore, no recommendation was provided in approximately a quarter of procedures. In 2012, the CQSC sent a letter with the European guideline recommendations to participating endoscopists, but the reminder letter had minimal impact on guideline adherence [5]. Although these results are disappointing, identifying that this problem exists is the first step to quality improvement in colonoscopy.

Fear of missing adenomas or CRC frequently drives this nonadherence [2] [7] [8]. When CRC is diagnosed before the patient is scheduled to return for surveillance colonoscopy, this is defined as interval CRC or “missed” CRC. Interval CRC may be the greatest fear of endoscopists, and many believe that shortening intervals will minimize this problem [2] [7] [8]. For example, if a patient has one small adenoma, the guidelines recommend repeat colonoscopy in 10 years. However, if the endoscopist worries that an adenoma could have been missed, repeat colonoscopy in 3 – 5 years may be recommended because that allows the endoscopist to remove a “missed” adenoma before it turns into CRC. There is one major problem with this approach: interval CRC usually occurs within 3 years of the index colonoscopy.

VA Cooperative Study 380 provides some of the best prospective data concerning this issue [9]. Average-risk veterans who underwent a screening colonoscopy between 1994 and 1997 were prospectively followed for up to 10 years. Veterans with 1 – 2 small adenomas at baseline or no adenomas at baseline “showed a dramatic decline in CRC incidence after 3 years.” In other words, the incidence of CRC was minimal in years 4 – 10 after colonoscopy. Most of the interval or “missed” CRC in low-risk patients occurred within the first 3 years. This is probably because most interval CRCs occur after endoscopists “miss” large flat sessile serrated adenoma/polyps or traditional serrated adenomas in the right side of the colon, which can rapidly grow into CRC [10]. Therefore, it is crucial to educate endoscopists that a 10-year interval is not too long in low-risk patients, provided a high quality colonoscopy has been performed.

How can endoscopists determine whether they routinely perform high quality colonoscopy with a low risk of interval CRC? Adenoma detection rate (ADR) is the key [2] [11]. ADR is defined as the incidence of adenomas among average-risk patients undergoing their first screening colonoscopy. Multi-society [2] and European [11] guidelines state that endoscopists should achieve an ADR of at least 25 %. Although this is the target, endoscopists should still strive to surpass this target. Corley et al. [12] demonstrated that the risk of interval CRC is lowest among endoscopists with ADRs of ≥ 33.5 %. Even if ADRs of 33.5 % or higher are not achieved, incremental gains in ADR > 25 % are still important. According to Corley et al. [12], every 1 % increase in ADR decreased the risk of an interval CRC by approximately 3 %.

Establishing other benchmarks or targets for high quality colonoscopy is crucial to improve adherence to guideline recommendations, and the ESGE recently completed this work [11]. In addition to a target ADR of 25 %, the ESGE now recommends that 95 % of post-polypectomy surveillance recommendations should adhere to guidelines. The target for adherence is now established and has been publicized. Research now needs to identify the causes of nonadherence in order to implement appropriate interventions.

Suboptimal or “fair” bowel preparation is one obvious cause of nonadherence. “Fair” bowel cleansing on the Aronchick scale is frequently reported when there is some liquid stool in the right side of the colon, but the bowel cleansing is not truly “inadequate.” These patients are most likely to be told to return sooner than recommended by guidelines [4] [7]. There is some substance to this rationale, as “fair” bowel preparation is associated with increased adenoma “miss” rates [7]. However, as discussed above, the solution is not to repeat colonoscopy in 3 – 5 years. Instead, if endoscopists are not satisfied with bowel cleansing and think there is a possibility that they could have missed a large flat serrated adenoma/polyp or traditional serrated adenoma, then the bowel preparation should be recorded as “poor” or “inadequate” and colonoscopy repeated within 12 months. This approach may seem problematic if many patients have a “fair” bowel preparation, but that should encourage endoscopists or endoscopy services to alter the bowel preparation protocols. The new ESGE benchmarks recommend that ≥ 90 % of bowel preparation should be adequate [11], which I equate to “good” or “excellent” cleansing on the Aronchick scale. This can usually be achieved with split-dose bowel preparation, which is the ESGE standard of care for bowel preparation [13]. If split-dose bowel preparation is already being used but with poor results, then the tools used to educate patients should be reassessed [13].

Finally, simply measuring guideline adherence [2] is an important quality improvement tool. The “Hawthorne effect” or “observer effect” consistently shows that an individual’s behavior changes when they know that their performance of a specific task is being closely observed and evaluated. Therefore, if endoscopists know that their guideline adherence is being monitored and that they are being compared with other endoscopists, then adherence rates are likely to increase. The Austrian CQSC system is ideal for this because it provides personalized biannual benchmark reports on quality indicators, along with comparisons to other participating endoscopists. This approach contributed to a substantial increase in ADR among participating endoscopists from 2007 to 2014 [6]. Adherence to guideline recommendations may also improve if these reports are modified to re-emphasize appropriate surveillance intervals and to educate endoscopists about new ESGE benchmarks.