Introduction
Individuals with high risk polyps have an elevated risk for colorectal cancer (CRC)
compared with those without high risk features [1]
[2]
[3]
[4]. Surveillance colonoscopy aims to reduce the risk of subsequent CRC for these individuals,
and involves a repeat colonoscopy after several months or 3 years, depending on the
polyp findings and polypectomy technique [5]
[6]
[7]. In 2020, the European Society of Gastrointestinal Endoscopy (ESGE) updated the
recommendations for these intervals, with high risk individuals defined as those with
≥5 adenomas, a serrated polyp/adenoma ≥10 mm, a serrated polyp with dysplasia, or
an adenoma with high grade dysplasia [8]. These individuals should receive a written recommendation for follow-up colonoscopy
in 3 years, while others can return to screening with colonoscopy in 10 years. Individuals
who undergo piecemeal resection of large polyps should have a repeat colonoscopy in
6 months. The main changes in the ESGE guideline update were the introduction of high
risk serrated polyps as a new entity requiring surveillance, and the exclusion of
villous histology as a high risk feature in adenomas. Additionally, the multiplicity
of adenomas requiring surveillance increased from 3 to 5 [9]. For large polyps removed by piecemeal resection, the cutoff for short-term follow-up
colonoscopy of 6 months was elevated from 10 mm to 20 mm [8].
Adherence to post-polypectomy guidelines is notoriously poor and is consistently reported
to be below the 95% target standard proposed by the ESGE in many countries [10]
[11]
[12]. Notably, interventions such as reminder letters sent to endoscopists are not sufficient
to significantly improve guideline adherence [13]. Over- and under-use of surveillance colonoscopy might have detrimental effects:
an interval that is too long might miss polyps that have recurred, and one that is
too short exposes individuals to the unnecessary harms associated with colonoscopy.
Nonadherence to interval recommendations adds to the colonoscopy demand and can impact
the cost-effectiveness of colonoscopy procedures and screening programs [14].
The introduction of the universal 10-mm cutoff for high risk polyps irrespective of
histology and removal of villous growth pattern as a high risk criterion has simplified
stratification. Little is known about whether this simplification has helped to improve
guideline adherence or about the temporal trends in guideline adherence by polyp type
since the introduction of the 2020 guideline update. The aim of this retrospective
analysis was to assess adherence to the 2020 post-polypectomy guideline intervals,
to reveal the trends in adherence before and after publication of the update, and
to investigate whether guideline interval adherence differed by polyp histology (conventional
adenoma or serrated polyp).
Methods
Study setting
Austria has an opportunistic primary colonoscopy screening program, where individuals
≥50 years irrespective of sex are eligible. Since 2007, the screening program has
been accompanied by a nationwide quality assurance program, the Austrian Quality Certificate
for Screening Colonoscopy. The aim of this voluntary program is to monitor screening
colonoscopy performance and to provide feedback for endoscopists by giving access
to biennial benchmarking reports. Participation in the program is not a requirement
for reimbursement of screening colonoscopy. Endoscopists are required to upload data
of screening colonoscopies to obtain certification; these include patient demographics,
colonoscopy findings, and data on colonoscopy quality parameters such as bowel preparation
and cecal intubation. The data are acquired through a standardized form, which can
be accessed through the endoscopists’ electronic health record or an online browser-based
application. The recommended follow-up interval in months is a non-mandatory field
in this form. The uploaded data are audited once a year, when endoscopists are asked
to share the written colonoscopy report of a random sample of the uploaded screening
colonoscopies. One random sample is drawn from colonoscopies where a polyp was detected,
and two random samples are drawn from negative colonoscopies. Only when data uploaded
to the database align with the original colonoscopy report is the data audit passed.
If endoscopists fail the audit, they are contacted and reminded of the program’s certification
standards.
The period of certification spans 2 years, and endoscopists are required to apply
for renewal of participation at the end of each cycle. To become re-certified, endoscopists
need to have passed the two data audits and have uploaded a sufficient number of colonoscopies
to ensure an adequate sample size for the benchmarking of key performance measures.
With the biennial invitation for re-certification, a letter is sent to each endoscopist
with information on current or previous guideline updates concerning screening colonoscopy.
In the invitation letter at the end of 2021 for the re-certification period of 2022–2023,
the guideline update for polypectomy, which was published by the ESGE in August 2020
and by the Austrian Society for Gastroenterology and Hepatology in December 2021,
was included. The post-polypectomy recommendations of the national society are in
line with the recommendations of the ESGE.
The study was approved by the Ethics Committee of the Medical University of Vienna
(EK 1095/2025).
Study population
Individuals undergoing screening colonoscopy and consenting to data transfer to the
Austrian Quality Certificate for Screening Colonoscopy were included. We excluded
screening participants who were diagnosed with CRC at colonoscopy, who had inadequate
bowel preparation (poor or insufficient bowel preparation on the Aronchick scale),
incomplete colonoscopies (as defined by not reaching the cecum), individuals without
polyps at screening, colonoscopy reports without a recommendation for follow-up colonoscopy
(either screening or surveillance recommendation), and those without complete polypectomy.
Incomplete polypectomy was defined as a colonoscopy where not all polyps were removed.
We also excluded individuals who received diagnoses other than colorectal polyps at
screening colonoscopy. These exclusion criteria were applied to ensure that only high
quality colonoscopies with complete polypectomy, to which the guidelines apply, were
used for the analyses.
Definition of variables and follow-up
Low risk screening participants
In line with the revised 2020 post-polypectomy ESGE guidelines, screening participants
who had low risk findings were defined as those with 1–4 adenomas, polyps <10 mm,
adenomas with low grade dysplasia, or serrated polyps without dysplasia.
High risk screening participants
High risk participants were defined as those with polyps ≥10 mm, an adenoma with high
grade dysplasia, or a serrated polyp with dysplasia. Those with ≥5 adenomas were also
considered high risk individuals, irrespective of the presence of high risk features
in polyps.
Correct colonoscopy interval recommendation
The correct colonoscopy interval recommendation was defined by the colonoscopy finding.
We considered a correct follow-up interval as “return to screening” for low risk findings
(i.e. a repeat colonoscopy after 10 years [120 months]). The correct interval for
high risk screening participants was defined as 36 months if high risk polyps were
sized 1–19 mm. If polyps ≥20 mm were detected at colonoscopy, the correct interval
was follow-up of 3–6 months. The ESGE also recommends this short interval for piecemeal
polypectomy of nonpedunculated polyps ≥20 mm. However, only the type of snare used
for polypectomy was recorded in the database, not whether piecemeal resection was
performed. Therefore, follow-up recommendations for sessile or flat polyps ≥20 mm
that underwent attempted complete polypectomy were defined as “correct” if 3–6 months
of follow-up were recorded by the endoscopist. Follow-up of pedunculated polyps ≥20
mm was correct if assigned to 36 months.
Statistical analysis
Descriptive statistics of polyp findings were applied for colonoscopies performed
after August 2020 by absolute and relative frequencies for categorical variables,
and means with SDs and medians with interquartile ranges (Q1, Q3) for continuous variables.
The relative frequency of correctly recommended surveillance intervals per year was
calculated as the total number of colonoscopies in which participants with high risk
polyps received the correct recommendation for high risk polyps (36 months) divided
by all colonoscopies performed in that year overall and subgrouped by endoscopy specialty.
A separate analysis of recommended follow-up was performed for low risk polyps and
polyps ≥20 mm. As incomplete histological removal of polyps might trigger shorter
follow-up interval recommendations, we performed a sensitivity analysis where the
correct recommendation rate for high risk polyps was calculated only for histologically
confirmed completely excised polyps. To investigate the time trends in surveillance
interval recommendations, we performed a piecewise linear regression with one breakpoint
where the dependent variable was frequency (%) of correct surveillance intervals per
month since August 2020. This model was chosen to identify a potential inflection
point in the recommendations over time. The appropriateness of model fit was assessed
by visualizing the model’s residuals. Additionally, we investigated whether a more
flexible model (spline transformation of time with two degrees of freedom) could yield
a better fit to the data (see Fig. 1s in the online-only Supplementary Material).
We were interested in the effect of the year of colonoscopy on the probability of
a correct surveillance interval. We fitted a mixed-effects logistic regression model
with the year of colonoscopy adjusted for age and sex of the participants and specialty
of the endoscopists. The estimates of this model are reported as odds ratios (OR)
with 95%CIs. We chose the endoscopists performing the colonoscopies as a random effect
to account for clustering between physicians. As there were only a few repeat colonoscopies
per patient, we did not add a random effect for the patient level.
We calculated the endoscopists adenoma detection rate (ADR) as the sum of all detected
adenomas up until the most recent colonoscopy divided by all colonoscopies performed
in a dynamic manner [15]. This way of calculating differs from interval-based ADRs (i.e. monthly or yearly
ADRs) in that they are updated at every single colonoscopy performed. The dynamic
ADR therefore allows for fluctuations over time, capturing its variability. Another
advantage of the dynamic ADR is that it only uses the data of detected adenomas up
until the current time point and does not rely on data points “ahead in time” (which
would be the case in interval-based ADR calculations) [15]. The dynamic ADR and the year of endoscopy were added to the regression model as
fixed effects. The model structure is depicted in Fig. 2s.
We performed a sensitivity analysis where polyps ≥20 mm were excluded, as there were
only limited data on the resection technique for these lesions. This could introduce
bias due to a variability in the recommended follow-up interval.
P < 0.05 was considered significant for hypothesis testing according to year of endoscopy;
the P value for trend was calculated with the mixed model. All analyses were performed with
R version 4.4.2 with the packages segmented version 2.1–3 and lme4 version 1.1–35.5
(R Foundation for Statistical Computing, Vienna, Austria).
Results
Baseline characteristics
Of 91 234 colonoscopies with polypectomy in which polyps were detected ([Fig. 1]), 59 052 were performed after 1 August 2020. Colonoscopies were performed by 342
endoscopists. Of these procedures, 9491 were performed in hospital endoscopy units,
18 624 by private practice surgeons, and 30 937 by private practice internal medicine
specialists. Polyp removal with either hot or cold snare (n = 30 741) and/or forceps
(n = 37 559) was performed. There were 6746 colonoscopies with high risk polyps requiring
surveillance, of which 6325 were screening colonoscopies and 421 were surveillance
colonoscopies. The median number of detected polyps was 2.0, and 9.1% of colonoscopies
with detected polyps reported a size of >10 mm ([Table 1]). Overall, the most advanced pathology was a serrated polyp in 22 579 colonoscopies,
and adenomas were detected in 36 473 colonoscopies.
Fig. 1 Flow chart of colonoscopies with polypectomy performed in 2018–2024. CRC, colorectal
cancer.
Table 1 Baseline characteristics of polypectomies performed after publication of the updated
guideline for post-polypectomy surveillance by the European Society of Gastrointestinal
Endoscopy (August 2020).
|
Variable
|
Colonoscopies1
(N = 59 052)
|
|
Q, quartile.
1Colonoscopies with polypectomy performed from August 2020 to November 2024.
|
|
Age
|
|
|
63.7 (8.62)
|
|
|
63.0 (57.0, 70.0)
|
|
Sex, n (%)
|
|
|
26 618 (45.1)
|
|
|
32 434 (54.9)
|
|
High risk polyp, n (%)
|
|
|
6746 (11.4)
|
|
|
52 306 (88.6)
|
|
Specialty, n (%)
|
|
|
9491 (16.1)
|
|
|
30 937 (52.4)
|
|
|
18 624 (31.5)
|
|
Polyp size (mm), n (%)
|
|
|
34 458 (58.4)
|
|
|
19 183 (32.5)
|
|
|
4553 (7.7)
|
|
|
847 (1.4)
|
|
|
11 (0.02)
|
|
Shape, n (%)
|
|
|
18 046 (30.6)
|
|
|
5013 (8.5)
|
|
|
35 982 (60.9)
|
|
|
11 (0.0)
|
|
Number of polyps
|
|
|
2.2 (1.9)
|
|
|
2.0 (1.0, 3.0)
|
|
|
2 (0.02)
|
|
Adenoma count
|
|
|
1.7 (1.3)
|
|
|
1.0 (1.0, 2.0)
|
|
|
16 994 (28.8)
|
|
Indication, n (%)
|
|
|
4157 (7.0)
|
|
|
54 895 (93.0)
|
Interval recommendations for polyps requiring surveillance
The frequency of correctly assigned surveillance intervals in high risk individuals
improved during the study period, with the strongest increase for adenomas ([Table 2]). In 2020, 38.4% of participants with high risk adenomas and 52.4% of participants
with high risk serrated polyps received the correct surveillance recommendation, which
increased to 51.2% and 54.3% in 2024, respectively (Table 1s). Overall, 41.5% of follow-up recommendations were correct in 2020, and this increased
after publication of the guideline update to 51.9% by November 2024, without a change
in the median length of the surveillance recommendation (36 months). In 2020, 11.2%
of surveillance interval recommendations were too long; this also increased to 15.4%
by November 2024. However, the frequency of surveillance intervals that were too short
decreased, from 47.4% in 2020 to 32.6% in 2024. There were 818 colonoscopies after
August 2020 with polyps ≥20 mm, for which a correct follow-up recommendation of 6
months was given to 30.1% of participants in 2020 and 33.3% in November 2024 (Table 2s). An incorrect interval length of >6 months was assigned for 32.9% in 2020, increasing
to 41.5% in 2024 for these polyps. Given that poorer bowel preparation might prompt
endoscopists to recommend shorter follow-up intervals, we performed a stratified analysis
for surveillance recommendations by bowel preparation degree (Table 3s). The proportion of correct surveillance interval recommendations was highest for
individuals with excellent bowel preparation compared with good or adequate bowel
preparation (Table 3s). The rate of correctly assigned surveillance intervals only marginally improved
when histologically confirmed incompletely excised polyps were removed from the analysis
(51.9% vs. 54.2%) (Table 4s).
Table 2 Median surveillance recommendations in months and their frequency in percent for high
risk polyps1 requiring surveillance according to the 2020 European Society of Gastrointestinal
Endoscopy guideline for post-polypectomy surveillance.
|
Median surveillance interval, months
(% of surveillance interval assignments)
|
|
2020
|
2021
|
2022
|
2023
|
2024
|
|
1High risk polyps are defined as adenomas ≥10 mm, adenomas with high grade dysplasia,
≥5 adenomas, serrated polyps ≥10 mm, or serrated polyps with dysplasia.
|
|
Adherence to guideline recommendation
|
|
|
36 (41.5)
|
36 (44.3)
|
36 (47.3)
|
36 (49.3)
|
36 (51.9)
|
|
|
60 (11.2)
|
60 (10.5)
|
60 (13.6)
|
60 (12.8)
|
60 (15.4)
|
|
|
12 (47.4)
|
12 (45.2)
|
12 (39.2)
|
12 (38.0)
|
12 (32.6)
|
Interval recommendations for polyps not requiring surveillance
Only 1.6% of screening participants with low risk polyps received the correct interval
recommendation for follow-up colonoscopy in 2020; this increased to 14.9% by November
2024. While 98.4% of low risk polyps were recommended intervals that were too short
in 2020 (median 36 months) (Table 3), this proportion decreased to 85.1% (median 60
months) ([Table 3]) by November 2024. After August 2020, there were 139 individuals with low risk polyps
who previously had high risk polyps removed at colonoscopy. The median interval recommendation
was 60 months in this group. In individuals with low risk polyps after previous removal
of polyps ≥20 mm (n = 22), the median recommended interval was also 60 months.
Table 3 Median surveillance recommendations in months and their frequency in percent for low
risk polyps1 not requiring surveillance according to the 2020 European Society of Gastrointestinal
Endoscopy guideline.
|
Median surveillance interval, months
(% of surveillance interval assignments)
|
|
2020
|
2021
|
2022
|
2023
|
2024
|
|
1Low risk polyps are defined as adenomas or serrated polyps <10 mm, serrated polyps
without dysplasia, adenomas with low grade dysplasia, or <5 adenomas.
|
|
Adherence to guideline recommendation
|
|
|
120 (1.6)
|
120 (4.0)
|
120 (7.3)
|
120 (11.3)
|
120 (14.9)
|
|
|
36 (98.4)
|
36 (95.9)
|
48 (92.7)
|
60 (88.7)
|
60 (85.1)
|
|
|
–
|
168 (0.1)
|
220 (0.1)
|
360 (0.1)
|
366 (0.1)
|
Temporal trends in surveillance interval recommendations
Before the introduction of the updated ESGE post-polypectomy surveillance guideline,
34.7% of endoscopists adhered to the 2013 guideline recommendations for advanced adenomas
in 2018; this rate remained stable throughout 2020 before publication of the updated
guideline (Table 5s). After publication of the updated guideline in August 2020, adherence to the recommended
intervals for high risk polyps gradually improved over time; however, the strongest
positive slope was observed in the first years after publication ([Fig. 2]). Each year after 2020 was associated with improved guideline adherence (P trend <0.001), with the largest estimate observed for the year 2024 (OR 1.49, 95%CI
1.19–1.86) ([Table 4]). In our sensitivity analysis, where polyps ≥20 mm were excluded, the estimates
for the year of endoscopy were comparable (Table 6s).
Fig. 2 Trends in guideline adherence over time by specialty.
Table 4 Mixed logistic regression model of the association of patient characteristics and
endoscopist characteristics with guideline adherence.
|
Characteristic
|
OR
|
95%CI
|
P value
|
P trend
|
|
ADR, adenoma detection rate; OR, odds ratio.
|
|
Year (vs. 2020)
|
<0.001
|
|
|
1.15
|
0.92–1.43
|
0.20
|
|
|
|
1.35
|
1.08–1.68
|
0.008
|
|
|
|
1.39
|
1.12–1.74
|
0.003
|
|
|
|
1.49
|
1.19–1.86
|
<0.001
|
|
|
ADR
|
1.01
|
1.00–1.03
|
0.06
|
|
|
Specialty (vs. hospital)
|
|
|
1.31
|
0.95–1.81
|
0.10
|
|
|
|
0.82
|
0.57–1.19
|
0.30
|
|
|
Sex (vs. female)
|
|
|
0.95
|
0.85–1.05
|
0.30
|
|
|
Age
|
1.00
|
0.99–1.00
|
0.12
|
|
Association of endoscopist characteristics with guideline adherence
We found no significant association of the endoscopist specialty (OR for surgery 0.82,
95%CI 0.57–1.19; OR for internal medicine 1.31, 95%CI 0.95–1.81) or setting other
than private practices compared with hospitals and adherence to the recommended guidelines
([Fig. 3]). However, there was a positive association of endoscopist’s ADR with the probability
of adhering to the ESGE guideline for surveillance of high risk polyps (OR 1.01, 95%CI
1.00–1.03).
Fig. 3 Trends in guideline recommendations since publication of the updated European Society
of Gastrointestinal Endoscopy guideline for post-polypectomy surveillance (August
2020).
Discussion
In the present study, we observed a general improvement in adherence to the recommended
guideline in a nationwide CRC screening program in Austria. However, the recommended
intervals were only correct in 51.9% of individuals with high risk findings and 14.9%
of individuals with low risk polyps at screening colonoscopy. Adherence to post-polypectomy
guidelines was found to be poor in many previous studies [11]
[16]
[17]
[18]
[19]
[20]. The reasons for nonadherence are not entirely understood. Endoscopists who provide
high quality colonoscopies and have a higher ADR are more likely to provide the correct
recommendations to patients [13]. In our study, we found that the ADR, as a measure of high performing endoscopists,
was positively associated with a higher probability of assigning a correct surveillance
interval. However, the effect size of the estimate was only small, implying that various
other factors contribute to the variability in surveillance recommendations.
A study from a tertiary academic center in the USA supports the notion that the endoscopists’
preferences rather than patient characteristics lead to incorrect interval recommendations.
Nonadherence to the United States Multi-Society Task Force on Colorectal Cancer post-polypectomy
guideline was more frequent for endoscopists who finished training more than 10 years
ago or those who had a higher annual colonoscopy volume, while endoscopist sub-specialty,
patient sex, or patient age showed no significant association [21]. Our data are in line with these findings, as we also found no significant association
of endoscopy specialty or participant characteristics with guideline nonadherence.
Another study from Italy found that high volume centers in the national screening
program and the provision of written recommendations were associated with correct
surveillance intervals [22]. Motives for shorter intervals than necessary might be attributable to overestimation
of the time to polyp recurrence risk, lack of knowledge of the guideline recommendations,
or financial interests [20]
[23]
[24]
[25].
Schoen et al. found poor adherence to post-polypectomy intervals in community practice,
and a high overuse of repeat colonoscopy in low risk individuals [25]. Our data from average risk individuals from a nationwide colonoscopy screening
program confirm the disproportionally low adherence in low risk participants compared
with high risk individuals, although the correct recommendations generally improved
during the 4-year study period.
The improvements in guideline adherence in our study from 2020 to 2024 were observed
in participants with polyps requiring surveillance (high risk findings) and polyps
not requiring surveillance (low risk findings), but the strongest improvements were
in the group of participants not requiring surveillance (1.6% in 2020 vs. 14.9% in
2024). Interestingly, the correct surveillance rate remained steady for serrated polyps
but strongly improved for adenomas during the study period. Two factors might have
contributed to this observation. On the one hand, there could be a rising awareness
that villous adenomas or adenoma multiplicity are no longer high risk features. On
the other hand, it has long been held that serrated polyps are innocuous lesions,
prompting endoscopists to recommend longer follow-up intervals compared with the follow-up
for adenomas [26]. Adherence to the surveillance intervals for advanced adenomas from the 2013 ESGE
guideline was poorer in the control period (January 2018–August 2020), which might
be explained by the more simplified guideline recommendations introduced in 2020 [8]
[9]. The ESGE recommendations for surveillance of serrated polyps remained mainly unchanged
in the guideline update of 2020. However, awareness of these kinds of polyps is still
poor, which is why the adherence to the guideline for high risk serrated polyps experienced
almost no improvement over time. This might be due to a lack of confidence in the
guideline recommendations; most of the incorrectly assigned intervals were too short
in this category. Although regular educational efforts such as seminars and courses
were provided by the project lead of our quality assurance program, serrated polyp
management remains only fair.
In general, we observed a reduction in surveillance overuse during the study period,
which mostly stemmed from a reduction in shorter intervals for low risk findings.
However, we also observed a small increase in surveillance recommendations that were
too long for high risk polyps (11.2% in 2020 and 15.4% in 2024). The correct surveillance
recommendations for polyps ≥20 mm remained relatively steady. The strongest improvements
in guideline adherence over time were observed in the first years after publication
of the update, but only significantly improved in the year 2024. A study from the
Netherlands addressed adherence to the 2020 ESGE post-polypectomy guideline. Among
a survey of 84 gastroenterologists, the authors found a median correct interval recommendation
in 71% of answers. In contrast to our study, the correct interval rate was lower for
serrated polyps [27].
Several measures can be taken to improve guideline adherence. Quality improvement
initiatives posting current guidelines where they are most visible (workstations,
pocket cards) can help physician awareness [28]. To overcome the issue of incorrectly assigned recommendations by endoscopists,
automatic reminders to primary care providers and patients can be distributed through
electronic health records [29]. However, an important and well-established practice to ensure that surveillance
is utilized is a written recommendation on the colonoscopy report [8]. Patients are often not aware of the need for surveillance after colonoscopy [30].
Our study had a large sample size of over 90 000 high quality colonoscopies with polypectomy
from a nationwide registry. Another strength of our study is that we analyzed the
trends in post-polypectomy guideline adherence spanning several years, rather than
adopting a cross-sectional design. Additionally, we were able to assess the temporal
trends in guideline adherence for adenomas and serrated polyps separately. We only
used high quality endoscopies, as recommendations for follow-up colonoscopy with poor
bowel preparation or failed cecal intubation would have skewed the data toward shorter
recommendations [31]. The data structure allowed for sensitivity analyses to identify potential biases
toward surveillance intervals that were shorter than recommended, such as histologically
confirmed incomplete removal of polyps or bowel preparation that was worse than “excellent.”
Limitations of this study include the lack of standardized polyp size evaluation,
which is
recommended by the ESGE. Polyp size measurement was left to the discretion of endoscopists.
The lack of data on the resection technique used for lesions ≥20 mm is a limitation;
only the
type of snare is recorded, but not whether piecemeal resection was attempted. This
information
is crucial for the distinction between a 3-year and a 3–6-month follow-up interval.
However,
piecemeal resection is the most common practice for flat/sessile lesions ≥20 mm in
Austria,
which is why all flat/sessile polyps were assumed to be removed by piecemeal resection.
Another limitation is that we cannot prove whether the electronically uploaded recommended
interval was also stated on the written colonoscopy report. Endoscopists enter data
personally
in the electronic report form, which should be derived from the written report they
provide
their patients. Therefore, we can only assume that that the uploaded interval is in
fact what
endoscopists advised their patients. Another limitation is that we could not analyze
the
within-center variability of interval recommendations, owing to the database structure.
In conclusion, when societies publish guidelines for post-polypectomy surveillance
intervals, a rise in adherence to the recommendations can be expected; in our study,
however, only small increments in guideline adherence were observed over a 4-year
period.
