Current evidence indicates that index screening and surveillance for colon cancer is effective [3,4,13]. However, specifics around appropriate and proper surveillance colonoscopy after identification of polyps is challenging and perhaps controversial [11,14] because recommendations should in theory balance the benefit of identifying early malignancy vs. harm engendered by the risks associated with colonoscopy [15]. For example, recommending inappropriate intervals for surveillance colonoscopies increases healthcare costs, increases the usage of medical resources, and decreases the capacity for patients who are in greater need of screening or surveillance colonoscopies [16,17]. Performing colonoscopy more frequently than needed also means an increased risk for complications such as bleeding or perforation [15–17].
Our data suggests that recommended surveillance colonoscopy intervals adhere more closely to the 2012 than the 2020 USMSTF guidelines for interval colonoscopy follow-up. Notably, recommended intervals for the 1st surveillance after index screening colonoscopy had a much lower adherence rate overall, regardless of the time frame for USMSTF guidelines. Finally, 2nd surveillance intervals were less closely followed than the 1st surveillance interval and were worse for the 2020 than the 2012 guidelines. Interestingly, adherence rates appeared to follow these trends, regardless of polyp features (i.e., whether LRA or HRA).
The 2020 guidelines had some important updates and changes compared to the 2012 guidelines. LRAs had the lowest adherence rate, which could be explained by the notable recommendation changes as follows: for 1 to 2 tubular adenomas < 10mm, the 2012 recommendation for follow-up was in 5–10 years, which was moved to 7–10 years with the new guidelines (Supplemental table 2). The dip in LRA compliance could potentially be explained by the change in the 2020 guidelines for the recommended interval for followup of 1–2 tubular adenomas (i.e., increased to 7–10 years; that is to say that with the new guidelines, the new minimum recommended interval had increased by 2 years. The new guidelines are also further stratified with follow-up recommendations for tubular adenomas after the 1st surveillance is performed. In contrast, the 2012 guidelines did not incorporate findings from previous colonoscopies in recommended intervals. With only “moderate” quality of evidence for follow-up of tubular adenomas, endoscopists may not be comfortable in recommending a longer waiting for the next surveillance colonoscopy [11]. This could be consistent with the stated “very low”, “weak”, or “moderate” evidence supporting some of the USMSTF guidelines [11]. In a study that surveyed physician opinions about 2012 polyp surveillance guidelines, 57% of gastroenterologists found the guidelines to be “very influential” in their practice. They also reported that although gastroenterologists were familiar with the guidelines, 76% disagreed with the recommendations [18].
Other theories of low compliance has been suggested, such as lack of guideline awareness, but a recent study suggested that this may not explain the low compliance [19]. A study showed that when endoscopists were given LRA and HRA surveys with clinical vignettes, they were able to answer them correctly despite their low compliance in clinical practice [19]. Again, endoscopists may feel the low strength of evidence may not be justifiable enough to follow it in practice. This study was limited as it only accounted for screening colonoscopies and did not examine surveillance exams and was only performed at one medical center and thus may not be generalizable. Previous studies have examined colonoscopy surveillance adherence. In a 2019 meta-analysis that examined rates of adherence to surveillance guidelines (2012) based on physician recommendations in 16 studies found that the appropriate adherence rate for colonoscopy surveillance adherence rate was 49%, with a range of 15–91% [12]. Some studies have proposed that (poor) quality of bowel preparation and concern about missed lesions or unclear histological findings of polyps, may possibly explain the shorter interval recommendations that have been observed [20,21].
We also have considered the possibility that the colonoscopies examined were during the COVID-19 pandemic, which may have affected follow-up recommendations. A systematic review showed that COVID-19 had an impact on the number of screening colonoscopies that performed during the pandemic [22], revealing that the number of colonoscopies decreased. Whether this could have had an effect on follow-up recommendations is unclear, though we speculate that the impact of COVID-19 on performance of colonoscopy would be unlikely to have a bearing on recommendations for future colonoscopies. Interestingly, consistent with our findings, a study that set out to determine whether the volume of colonoscopy could be reasonably reduced by more rigorously implementing 2020 USPSTF guidelines to help alleviate the downstream effort of COVID-19, found that 15–21% of colonoscopies at their institution qualified to be rescheduled to a future year [23].
We recognize strengths and limitations of our study. A critical strength of the study was that we performed a power analysis a priori to ensure an adequate cohort of patients in each of the 2012 and 2020 cohorts. We also examined adherence according to adenoma features (i.e., LRA and HRA), and we took care to exclude patients with inadequate preparations or pathology reports. In terms of weaknesses, perhaps the most important weakness was that this study was performed at a single center, and the practice at one institution may be different than at other institutions. However, we speculate that given current national healthcare policies, the practice at our institution is likely similar to others. We did not include patients in whom surveillance was recommended past the age of 75 and therefore cannot comment on practices in this age group. It should be emphasized that this was intentional since guidelines for surveillance past the age of 75 are controversial, and guidance is limited [24]. Further, the US Preventive Services Task Force recommends CRC screenings until age 75 and recommends individualized screening decisions for adults aged 76 to 85 years of age [5,24].
In conclusion, our findings indicate that adherence to surveillance guideline recommendations after index screening was better than surveillance colonoscopy and that adherence to surveillance guideline recommendations was especially poor after 2nd surveillance exams. Further, adherence has declined after the introduction of 2020 guidelines, even while it has been published for the past 2–3 years. We speculate the 2012 USMSTF guidelines may be easier to follow than the updated guidelines, but that further education and perhaps stronger evidence may help improved adherence to 2020 USMSTF guidelines.