The aim of this study was to explore the safety of colonoscopy surveillance for LS patients. The results of our study, for both SAEs and AEs, suggest an overall low complication rate, which are consistent with anticipates outcomes and comparable to those reported in similar studies.

Within the scope of this paper: AEs have been categorised as abdominal pain/discomfort or infectiously caused gastrointestinal symptoms e.g. loose stools or diarrhea. SAEs have been categorised as bleeding or perforation. The risk of bleeding for an LS patient per-examination in a surveillance programme, in comparison with a non-LS patient, is lower with a relative risk reduction of 0.06%, compared with results from Forsberg et al. (0.17%). There was no perforation (0%) in this study in comparison with Forsberg et al. (0.11%).

The rate of both AEs and SAEs for each LS patient are significantly higher. Over a life-time, the cumulative risk of complications for each LS patient is likely to be higher than for populations not enrolled in colonoscopy surveillance programmes.

As demonstrated, the majority of the complications occurred after 2011. Most presumably, this reflects better colonoscopy registrations for subsequent years rather than inferiorly performed colonoscopies. However, it cannot be neglected, that the colonoscopy technique has been improved during the timeframe of this study. For instance, the use of carbon dioxide instead of air for colonoscopy insufflation, an increasingly dominant method since the turn of the millennium, has demonstrated its advantages in regards to e.g. abdominal discomfort [22, 23]. It is noteworthy, as seen in Fig. 1, that the complications registered from 2016 to 2020 corresponds to approximately 1/100 in regards to colonoscopy-associated complications per-examination. From a Swedish context, there are unfortunately deficiencies in reporting, such as desaturation during examination. To provide a relatively fair picture, the decision was therefore made to exclude cardiovascular complications.

There are currently classifications to facilitate the characterisation of complications. The Classification for Adverse Events Gastrointestinal Endoscopy (AGREE) classification, provides a novel adapted tool specifically designed for endoscopy [24], although not widely used in clinical practice. As a result of still limited use, the decision was made not to use the AGREE classification.

Colonoscopy surveillance for LS patients aims to detect and remove potential precancerous adenomas, based on the assumption that surveillance-programmes are patient-safe, feasible and have a significant impact on CRC-mortality. In other words, the potential benefits of colonoscopy surveillance for LS patients must outweigh the risks. Due to the low average number of lifetime colonoscopies, the cumulative complication rate experienced by the overall population is low.

LS patients on the other hand, as a cohort, requiring recurrent annual or biennial colonoscopies, have an increased cumulative risk of colonoscopy-associated complications. Since LS patients are younger, healthier, and, hence, less prone to other comorbidities, the colonoscopy-associated complication rates become increasingly important. AEs and perforations, in particular, are of specific interest as they are affiliated with higher mortality- and morbidity rates [11].

For LS patients, colonoscopy-associated complications have long-term consequences beyond the immediate mortality and morbidity. These complications include reduced trust in the provided care, fear of future colonoscopies, and a reduced propensity to follow surveillance programmes with consequent risk of reduced adherence. Of note, LS patients’ adherence to surveillance programmes has been reported to be inadequate, partly due to patients’ perceived challenges associated with colonoscopy, including discomfort, embarrassment and the time-consuming nature of the procedure [25,26,27]. Patients experience bowel preparation as burdensome [28], particularly as LS patients must undergo bowel preparation repeatedly [29]. The first colonoscopy experience may determine the patient’s adherence to the surveillance programme [23]. An increased frequency of hospitalisation and follow-ups also increase health economic impacts.

It is well-known that the risk of the development of CRC differs between LS-genotypes. Other factors affecting the risk of CRC include sex, BMI, smoking history and non-participation in regular colonoscopy surveillance that contributes to the risk of CRC [19, 30, 31].

Historically, former guidelines on colonoscopy surveillance for LS patients in Sweden and Europe did not consider the LS-genotype when designing surveillance programmes [32]. This implies, that CRC-risk is unevenly distributed, but the colonoscopy-related complications are not, which is troublesome from a benefit–risk perspective [33]. Consequently, the colonoscopy surveillance programme should be targeted and individualised, and based on certain preconditions. However, on the basis of surveillance programmes’ arrangement, only a fraction of the surveillance programme’s participants will benefit. Moreover, this possible presumed benefit may vary for each participant into only modest longevity.

To our best knowledge, this is the first study of its kind, providing an assessment of colonoscopy surveillance-associated complications and colonoscopy safety in a Swedish setting of a large cohort consisting of 351 LS patients over 33 years of age. LS, as previously described, is a rare condition which presents challenges in the gathering of larger LS cohorts. Although LS patients represent a small subgroup of those undergoing colonoscopy, repeated colonoscopies increase individual LS patients’ lifetime risk, highlighting the importance of these data. Furthermore, the overall high-quality population data in Sweden regarding colonoscopy complications partially compensates for the lack of a direct comparison group of the study.

There are several limitations to this study. First, it is unknown to what extent the number of referred LS patients to Karolinska University Hospital corresponds with the actual number of LS patients in the Stockholm region. Hence, it is possible that there are colonoscopies as well as colonoscopy-associated complications not captured in the study. Moreover, there may be differences between the patients referred and not referred. Second, data were acquired from only one region in Sweden with mainly White individuals, which may decrease generalisability. Third, like any other retrospective cohort study, there may be limited- or missing data and, accordingly, there is risk of less accurate entry of data which may affect overall data quality.

Future studies should address whether individualised screening programmes affect the relative and absolute risk of surveillance-related colonoscopy complication for LS patients. In addition, studies comparing complication rates between centres with a high volume compared with a low volume of LS patient surveillance colonoscopies, would determine whether centralisation of surveillance programmes would impact complication rates.

In conclusion, this study demonstrates that per-examination colonoscopy complication rates in LS patients are comparable to those previously reported in the general population. However, understanding the lifetime risk of surveillance-related colonoscopy complications is important when designing targeted surveillance programmes, as LS patients require lifelong surveillance. The use of more specific and targeted programmes could increase cost-effectiveness and further positively affect the risk–benefit balance.