Colorectal cancer (CRC) is the third most common cancer globally, accounting for more than 1.8 million new cancer diagnoses per year (1). Most CRCs slowly develop over many years from adenomas which can be detected and removed by endoscopic screening. The 2 main options for endoscopic screening are flexible sigmoidoscopy (FS), which enables visualization of the distal colon and rectum, where most CRCs are located, and colonoscopy, which enables visualization of the entire colon and rectum. However, the more complete visualization by screening colonoscopy comes at the prize of the need of complete bowel cleansing, a major obstacle to screening adherence, substantially higher costs, and higher complication rates. A crucial question for choosing between colonoscopy and FS for CRC screening is therefore if and to what extent this prize is justified by colonoscopies' ability to additionally visualize the proximal colon and to prevent cancers in the proximal colon.
Multiple randomized controlled trials have demonstrated reduction of distal CRC incidence and mortality by FS-based screening (2–5). Randomized controlled trial results on long-term effects of screening colonoscopy will become available only many years from now (6), and evidence from prospective cohort studies keeps being very limited. A number of case-control studies have consistently suggested that screening colonoscopy was associated with strongly reduced CRC incidence and mortality (7–12), and they typically found stronger protection from cancer in the distal colon and rectum than from proximal colon cancer (13). However, apart from availability of limited covariate data in some of the studies, case-control studies on screening are prone to specific potential biases (14). The few prospective cohort studies were either very limited in sample size (15) or follow-up time (16) and measured colonoscopy exposure just once, i.e., at recruitment only (15,16), or focused on specific professional groups (such as female teachers or health care professionals) or age groups only (17–19) and on CRC incidence (19) or mortality only (18).
Germany was one of the first countries to offer colonoscopy as a primary CRC screening examination nationwide. Since October 2002, men and women aged 55 years or older have been entitled to have up to 2 screening colonoscopies 10 or more years apart (20). Certification to conduct screening colonoscopy is tightly regulated on the basis of extensive previous training, and its maintenance is subject to rigorous quality control. In particular, only physicians with specific qualifications (gastroenterologists, colorectal surgeons, or internists with colonoscopy board certification) and extensive experience (at least 200 colonoscopies and 50 polypectomies within the past 2 years) are eligible to perform screening colonoscopy. Completeness of colonoscopy as well as any finding during colonoscopy has to be photodocumented. Only colonoscopies reaching the cecum are reimbursed, with the exception of cases with stenosing tumors.
In this study, we aimed to assess the association between use of screening colonoscopy and CRC incidence and mortality in a prospective population-based cohort study of older adults from Germany, paying particular attention to specific effects of preventing cancer in the proximal and distal colon and rectum.
Study design and study population
Our analysis is based on data from the ESTHER study, an ongoing statewide population-based cohort study among older adults conducted in Saarland, Germany. Details of the study design have been reported elsewhere (21,22). Briefly, 9,949 male and female residents of Saarland aged 50–75 years with sufficient knowledge of the German language were recruited in 2000–2002 by their general practitioners (GPs) during a routine health check-up aiming at early detection of cardiovascular diseases and diabetes. The study population has been shown to closely resemble the study population of a representative German national health survey within the corresponding age range performed in 1998 with respect to major sociodemographic and health-related characteristics (23). ESTHER participants and their GPs are regularly recontacted every 2–3 years, and participants are followed up with respect to incidence and mortality of major chronic diseases including cancer. The study was approved by ethics committees of the Medical Faculty Heidelberg of Heidelberg University and of the Physicians' board of Saarland. Written informed consent was obtained from each participant.
For the current analysis, we excluded participants with missing information on screening colonoscopy before recruitment (n = 742). For analyses on CRC incidence, we additionally excluded participants with a CRC diagnosis before recruitment (n = 111), leaving a total of 9,096 participants for analysis of CRC incidence and 9,207 participants for analysis of CRC mortality.
At baseline, comprehensive information on lifestyle factors and medical history was obtained by self-administered standardized questionnaire from the participants, which was complemented by GP information from the available health check-ups and medical records. In particular, participants were asked if they ever had a colonoscopy for screening purposes, and if so, the date of the most recent examination. Follow-up information on screening colonoscopy was obtained by standardized questionnaires at 2-, 5-, 8-, and 17-year follow-up, conducted in 2002–2004, 2005–2007, 2008–2010, and 2017–2019, respectively. In the few cases where the date of screening colonoscopy was missing and could not be derived from medical records (3.6% in the 2-, 5-, and 8-year follow-up questionnaires and 0.7% in the 17-year follow-up questionnaire), the date was set at the midpoint of the respective time interval since the preceding follow-up round except for 4 cases who had a CRC diagnosis during this time interval for whom the date of screening colonoscopy was assumed to equal the date of CRC diagnosis.
Follow-up with respect to CRC incidence by the end of 2018 was conducted through record linkage with the statewide population-based Saarland Cancer Registry (SCR). The SCR collects any incident and fatal cancers of people residing in Saarland, irrespective of place of diagnosis. If individuals in the cohort had a cancer diagnosed outside Saarland, the cancer notification would be forwarded to and recorded in the SCR. Vital status by the end of 2018 and date of death could be ascertained by record linkage with population registries for 99.7% of the cohort. Information on cause of death could be obtained from 98.9% of deceased participants from public health authorities. Hence, analysis of CRC incidence included CRC-free person-times under observation and cases diagnosed with CRC (International Classification of Diseases, 10th Revision [ICD-10] codes C18-C20), and analysis of CRC mortality included person-times and deaths from CRC (ICD-10 codes C18-C20) until the end of 2018.
Validation by medical records was performed for 990 self-reported screening colonoscopies at the 5-year follow-up (2005–2007). Performance of colonoscopy was confirmed in 100% of cases, with completion rates (cecum reached) of 97%. Screening was confirmed as primary indication for 77% of colonoscopies. The polyp detection rate was 45% for men and 30% for women.
We first described the study population with respect to sociodemographic characteristics and other known or suspected CRC risk factors, including age, sex, school education (≤9, 10–11, and ≥12 years of schooling), history of CRC in a first-degree relative, smoking (never, former, or current), alcohol consumption (women: none, <20, or ≥20 g/d; men: none, <40, or ≥40 g/d), body mass index (<25, 25–29.9, and ≥30 kg/m2), physical activity (<1, 1–2, or ≥2 hr/wk of vigorous physical activity), red and processed meat consumption (≤1 time per week, multiple times per week, or ≥1 time per day), use of hormone replacement therapy (never, former, or current; women only), and physician-diagnosed diabetes.
Exposure status was initially defined at baseline and updated at 2-, 5-, 8-, and 17-year follow-ups. Baseline characteristics of participants who used screening colonoscopy and participants who never had screening colonoscopy before or after study enrollment were compared using χ2 tests for binary variables and Cochran-Armitage test for trend for multilevel variables. Cox proportional hazards models with screening colonoscopy as time-varying exposure variable (in that participants switched from unexposed to exposed when they had a screening colonoscopy during follow-up) were used to assess associations with CRC incidence and mortality, accounting for the above-mentioned covariates. We included person-time from start of enrollment to CRC incidence, death, or end of 2018 (whatever came first) for analyses of CRC incidence and from start of enrollment to CRC death, death from other cause, or end of 2018 (whatever came first) for analyses of CRC mortality. The proportional hazards assumption was evaluated and accounted for where necessary by interaction terms between covariates and time since enrollment. Associations with CRC incidence and mortality were assessed for ever versus never use of screening colonoscopy and, in addition, for use of screening colonoscopy within the preceding 10 years versus never use.
In all analyses, 2 types of models were run. Model 1 adjusted for sex and age only, and model 2 adjusted for all of the above-listed covariates that were associated with ever use of screening colonoscopy. Multiple imputation by chained equations (24) was applied to deal with missing values in the following covariates (missing values in parentheses): school education (n = 212, 2.3%), history of CRC in a first-degree relative (n = 105, 1.1%), smoking (n = 226, 2.5%), alcohol consumption (n = 825, 9.0%), body mass index (n = 15, 0.2%), physical activity (n = 26, 0.3%), red meat consumption (n = 475, 5.2%), processed meat consumption (n = 429, 4.7%), use of hormone replacement therapy (among women, n = 1, 0.02%), and diabetes (n = 131, 1.4%). The imputation procedure was applied under the assumption of data missing at random and 5 data sets were imputed using the variables included in the fully adjusted Cox regression models. Associations between screening colonoscopy and CRC incidence and mortality were quantified by hazard ratios (HRs) and their 95% confidence intervals (CIs).
Separate models were run for total, proximal, and distal CRC incidence and mortality. CRCs were defined as proximal if they were located proximal of the splenic flexure (ICD-10 codes: C18.0-C18.4) and as distal otherwise (ICD-10 codes: C18.5-C18.7, C19, and C20). Cases of overlapping neoplasms of the colon (C18.8) and unspecific neoplasms of the colon (C18.9) were not considered in site-specific analyses.
HRs by CRC subsites were estimated using a competing risk method where cases from the complementary site were censored at the date of diagnosis. We used the Wald test to compare HRs between sites (i.e., proximal vs distal). Sex-specific analyses on the association of screening colonoscopy with total and site-specific CRC incidence were conducted in addition to analyses for the whole study population.
Furthermore, to explore net effects of screening colonoscopy on total cancer incidence and mortality and to evaluate potential residual confounding related to health consciousness, we also quantified the association of screening colonoscopy with overall and non-CRC cancer incidence and mortality.
All statistical analyses were performed by the software package SAS (version 9.4; SAS Institute Cary, NC) and R version 3.6.2 (25). Two-sided P values <0.05 were considered statistically significant.
Table 1 shows characteristics of the study participants. Mean age at recruitment was 61 years, and 54.8% were women. At least 1 follow-up questionnaire was obtained from 96% of participants. Nonparticipation in follow-up surveys was higher among smokers and people with diabetes than nonsmokers and people without diabetes. By the 17-year follow-up, 5,388 participants (58.5%) had reported to have ever undergone a screening colonoscopy. Approximately 9% of participants had a first-degree relative with CRC, and these participants were more likely to have had a screening colonoscopy (68.2%) than participants without a family history (57.8%, P < 0.001). Other factors associated with use of screening colonoscopy were school education, smoking, alcohol consumption, body mass index, physical activity, red and processed meat consumption, use of hormone replacement therapy, and diabetes.
Follow-up of cancer incidence and mortality was complete for 99.8% and 99.7% of the cohort. During a median follow-up over 17.2 years, 268 incident CRC cases and 98 CRC deaths were observed. Table 2 shows the associations of screening colonoscopy with total, proximal, and distal CRC incidence and mortality estimated by Cox proportional hazards models. In age- and sex-adjusted analysis, a history of screening colonoscopy was associated with strongly reduced total CRC incidence and mortality, and these associations were essentially unchanged after controlling for additional potential confounders (adjusted HR [aHR] 0.44, 95% CI 0.33–0.57 for incidence and 0.34, 95% CI 0.21–0.53 for mortality). In site-specific analyses, stronger reduction was seen for distal (aHRs 0.36, 95% CI 0.25–0.51, and 0.33, 95% CI 0.19–0.59, respectively) than for proximal cancer (aHRs 0.69, 95% CI 0.42–1.13, and 0.62, 95% CI 0.26–1.45, respectively; P value for heterogeneity in incidence by cancer subsite = 0.002). Nevertheless, strong reduction of mortality from proximal cancer was also observed within 10 years after screening colonoscopy (aHR 0.31, 95% CI 0.10–0.96).
Strong reduction of distal CRC incidence but modest reduction of incidence of cancers in the proximal colon was consistently seen among both men and women (Table 3). Associations for distal CRC incidence were tentatively stronger for men than for women, but 95% CIs were overlapping.
As can be seen from Table 4, reduction of cancer incidence and mortality was exclusively seen for the colon and rectum, in particular for the distal colon and rectum, with consistent null associations for any other cancers. Nevertheless, because of the strong inverse association of screening colonoscopy with CRC incidence and mortality, screening colonoscopy was associated with a weak, nonsignificant reduction of overall cancer incidence and a significant reduction of overall cancer mortality (aHRs 0.92, 95% CI 0.84–1.01 and 0.77, 95% CI 0.66–0.89, respectively).
In this prospective population-based cohort of older adults from Germany, incidence and mortality from total and distal CRC was strongly reduced among participants who had undergone screening colonoscopy. Reduction of incidence and mortality was particularly strong for cancer in the distal colon and rectum, but strong reduction of mortality from proximal cancer was also observed within 10 years after screening colonoscopy.
To the best of our knowledge, this is the first cohort study with long-term follow-up reporting on the effects of screening colonoscopy on CRC incidence and mortality from Germany, one of the first countries with a nationwide offer of colonoscopy as primary screening examination. Our results are in line with and expand findings of a cross-sectional study among participants of screening colonoscopy in Germany, in which previous colonoscopy was associated with a strongly reduced prevalence of advanced neoplasms in the distal colon and rectum, but not in the proximal colon (26), and of a population-based case-control study from Germany that also found a strong risk reduction for distal CRC within 10 years after screening colonoscopy (9). The latter study had found a statistically significant reduction of risk of cancer in the proximal colon, albeit less pronounced. No results on cancer mortality were available from these studies.
To the best of our knowledge, only 1 prospective cohort study each has assessed the impact of screening colonoscopy on site-specific CRC incidence and mortality, respectively (17,18). Our results on site-specific CRC incidence are remarkably consistent with those from a cohort study among female teachers in France with 15 years of follow-up (17), which reported strong, statistically significant reduction of CRC incidence in the distal colon and rectum (aHRs 0.57 and 0.37, respectively) but not in the proximal colon (aHR 0.87). A cohort study among female and male health professionals with 22 years of follow-up (18) had also found a protective effect of screening colonoscopy against mortality from cancer in the proximal colon, which was although weaker than the protective effect against distal CRC (aHRs 0.47 and 0.18, respectively). Taken together, these results support suggestions that colonoscopy strongly protects from CRC overall by detecting and removing precursors of the disease. However, such protection does not seem to be uniform for cancers arising in different areas of the colon, with less protection for the proximal colon whose precursors are known to differ in many respects from precursors of distal cancers.
The different effects on proximal and distal CRC incidence and mortality may reflect different routes of carcinogenesis. In particular, proximal cancers more often develop from serrated polyps whose detection and removal pose major challenges to endoscopists (27,28). In addition, adenoma miss rates have also been reported to be higher in the proximal colon than in the distal colon and rectum (28). Although screening colonoscopy was not significantly associated with reduction of incidence and mortality in the proximal colon, we still found a trend toward reduced incidence of proximal cancer after screening colonoscopy. Furthermore, mortality from proximal cancer was also strongly and significantly reduced within 10 years from screening colonoscopy, possibly as a result of early detection of preclinical proximal cancer which is expected to result in higher cure rates, given the strong dependency of CRC survival from stage at diagnosis. However, sample sizes of site-specific analyses were still limited, which led to rather wide CIs around the site-specific effect estimates.
Stronger effects of screening colonoscopy in preventing distal CRC compared with proximal CRC suggested by our study are also in line with and may explain the observation that postcolonoscopy CRCs are much more frequently located in the proximal colon and more often demonstrate microsatellite instability than cancers detected at screening colonoscopy (29). A recent study suggested that risk reduction after colonoscopy was weaker for molecular subtypes characteristics of the proximal colon (i.e. microsatellite instability and CRC arising from the sessile serrated pathway) compared with their paired subtypes (i.e., microsatellite stability and CRC arising from the traditional pathway) (30). These patterns support suggestions that different molecular features of proximal cancers may contribute to the lack of effectiveness of colonoscopy in preventing them, besides the more difficult detection of precancerous lesions in the proximal colon.
Furthermore, a previous study has shown an overall sustained improvement in adenoma detection rate from 2003 through 2012 by screening colonoscopy in Germany, which might reflect favorable development and further optimization in colonoscopy performance across periods (20,31,32). Our findings of weaker protective effects for the proximal colon might thus partly reflect poorer quality of earlier colonoscopies.
Like previous studies, our study may underestimate true screening colonoscopy effects (in both the proximal or the distal colon and rectum) to some extent because information on colonoscopies conducted for diagnostic purposes, which are expected to convey similar protection by detecting and removing colorectal neoplasms, was not available. According to national health survey data from Germany, a substantial proportion of older adults also has had diagnostic colonoscopy (33). This proportion would be expected to be even higher among those not having had screening colonoscopy, as a screening colonoscopy might not be warranted in people who have had a recent diagnostic colonoscopy. The expected “contamination” by diagnostic colonoscopies of the reference group with no screening colonoscopy bears the potential of underestimation of the effects of screening colonoscopy (34). Screening colonoscopy may therefore have led to somewhat stronger reduction of total and site-specific CRC incidence and mortality than estimated by our study.
The strong effects of screening colonoscopy on total and distal CRC incidence and mortality estimated in this study are in agreement with recent trends of a major decrease in CRC incidence and mortality observed in Germany after nationwide introduction of the offer of screening colonoscopy in 2002 (35). Like in the United States, where widespread use of screening colonoscopy already started in the 1990s and an even stronger decline in CRC incidence and mortality has been observed since then (36), this decline is most pronounced at older ages covered by CRC screening (35,37). However, CRC screening still remains underused, and even stronger and more rapid declines of CRC incidence and mortality could be achieved by enhanced adherence to screening offers (38), especially among the high-risk groups with unfavorable risk factor profiles, such as obese people and smokers (39).
A major reason for nonadherence with offers of screening colonoscopy is the invasive nature of this examination, requiring complete bowel cleansing. Furthermore, costs, complication rates, and capacities needed are substantially higher compared with flexible sigmoidoscopy. The findings of a lower use of screening colonoscopy for reducing cancer incidence in the proximal colon support suggestions that even larger effects might be achieved by offers of flexible sigmoidoscopy, which might be used by larger proportions of the population, in particular if combined with fecal immunochemical testing (which enables detection of most proximal cancers) (40). Cost-effectiveness analyses have commonly assumed substantially stronger total CRC incidence reduction by 10-yearly colonoscopy compared with 5-yearly flexible sigmoidoscopy (41). Our results suggest that the advantage of screening colonoscopy over flexible sigmoidoscopy may have been overestimated. These suggestions are in line with those of a recent modeling study which suggested similarly strong effects of a single flexible sigmoidoscopy or a single colonoscopy in reducing CRC incidence and mortality (42). Our findings may therefore have important implications for refining analyses of effectiveness and cost-effectiveness of various screening offers and for further development and implementation of screening programs, in particular in countries with limited colonoscopy capacities.
In the interpretation of our study, a number of strengths and limitations deserve careful consideration. Strengths include long-term follow-up of a large cohort from the general population. Repeated follow-up examinations enabled exposure updates throughout follow-up, and comprehensive data collection enabled careful adjustment for relevant potential confounders. A major limitation is the lack of information on diagnostic colonoscopies. Furthermore, despite the overall large size of the cohort, numbers of CRC cases and deaths were still rather small, leading to wide CIs for some of the risk estimates and hindering further analyses for specific high-risk subgroups, such as people with a family history of CRC. Although we carefully controlled for multiple potential confounders, we cannot rule out residual confounding by unmeasured confounders. However, the selectivity of effects seen for total and distal CRC and not for any other cancer suggests potential residual confounding (e.g., by factors related to general health consciousness), if any, to be small. To the best of our knowledge, no previous observational study has included such a “selectivity check” of associations.
Despite its limitations, this prospective population-based cohort study adds important evidence on the effects of screening colonoscopy in reducing overall and site-specific CRC incidence and mortality, as well as total cancer mortality. Our results underline the large potential of screening colonoscopy to prevent cancer in the distal colon and rectum and to reduce mortality from these cancers and even total cancer mortality. At the same time, however, they also underline the need of further efforts toward more effective prevention of cancer in the proximal colon.
CONFLICTS OF INTEREST
Guarantor of the article: Hermann Brenner, MD, MPH.
Specific author contributions: H.B.: designed, led, and supervised the study and drafted and revised the article. F.G.: conducted the statistical analyses and revised the article. C.C.: contributed to the statistical analyses. B.S. and B.H.: contributed to the coordination and conduction of data collection and work-up of data. All authors contributed to the interpretation of the data, critically reviewed manuscript drafts, constructively contributed to their finalization, and agreed with the final version submitted. The researchers are independent from funders. All authors had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis.
Financial support: The ESTHER study was funded by the Baden-Württemberg State Ministry of Science, Research and Arts (Stuttgart, Germany), the German Federal Ministry of Education and Research (Berlin, Germany), the German Federal Ministry of Family, Senior Citizens, Women, and Youth (Berlin, Germany), and the Saarland State Ministry of Social Affairs, Health, Women, and Family (Saarbrücken, Germany). The analyses for this project were supported by a grant from the German Cancer Aid (No. 70112095). The funders had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Potential competing interests: None to report.
WHAT IS KNOWN
- ✓ Multiple randomized controlled trials have demonstrated that screening with flexible sigmoidoscopy can substantially reduce incidence and mortality from cancer in the distal colon and rectum.
- ✓ Evidence on the impact of screening colonoscopy on colorectal cancer incidence and mortality from randomized trials is lacking, and evidence from prospective cohort studies is very limited.
- ✓ In particular, it is highly uncertain to what extent screening colonoscopy can additionally reduce incidence and mortality from cancer in the proximal colon.
WHAT IS NEW HERE
- ✓ This population-based, prospective statewide cohort study from Saarland/Germany with repeat updates of exposure information demonstrates major reduction in colorectal cancer incidence and mortality among people who underwent screening colonoscopy.
- ✓ Reduction of incidence and mortality was much stronger for cancer in the distal colon and rectum than for cancer in the proximal colon, but strong reduction of proximal cancer mortality was also seen within 10 years after screening colonoscopy.
- ✓ These results may help to refine expectations of incremental effectiveness of colonoscopy screening over screening by flexible sigmoidoscopy in preventing colorectal cancer.
The authors gratefully acknowledge valuable contributions to the conduction of the ESTHER study by Hartwig Ziegler, Christa Stegmaier, Sonja Wolf, Gregor Thal, Martina Mohr, Jennifer Peter, Volker Herrmann, and Utz Benscheid.
1. Bray F, Ferlay J, Soerjomataram I, et al. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68(6):394–424.
2. Segnan N, Armaroli P, Bonelli L, et al. Once-only sigmoidoscopy in colorectal cancer screening: Follow-up findings of the Italian Randomized Controlled Trial—SCORE. J Natl Cancer Inst 2011;103(17):1310–22.
3. Holme O, Loberg M, Kalager M, et al. Effect of flexible sigmoidoscopy screening on colorectal cancer incidence and mortality: A randomized clinical trial. JAMA 2014;312(6):606–15.
4. Atkin W, Wooldrage K, Parkin DM, et al. Long term effects of once-only flexible sigmoidoscopy screening after 17 years of follow-up: The UK flexible sigmoidoscopy screening randomised controlled trial. Lancet 2017;389(10076):1299–311.
5. Miller EA, Pinsky PF, Schoen RE, et al. Effect of flexible sigmoidoscopy screening on colorectal cancer incidence and mortality: Long-term follow-up of the randomised US PLCO cancer screening trial. Lancet Gastroenterol Hepatol 2019;4:101–10.
6. Kaminski MF, Bretthauer M, Zauber AG, et al. The NordICC study: Rationale and design of a randomized trial on colonoscopy screening for colorectal cancer. Endoscopy 2012;44(7):695–702.
7. Cotterchio M, Manno M, Klar N, et al. Colorectal screening is associated with reduced colorectal cancer risk: A case-control study within the population-based Ontario Familial Colorectal Cancer Registry. Cancer Causes Control 2005;16(7):865–75.
8. Doubeni CA, Weinman S, Adams K, et al. Screening colonoscopy and risk for incident late-stage colorectal cancer diagnosis in average-risk adults. Ann Intern Med 2013;158:312–20.
9. Brenner H, Chang-Claude J, Jansen L, et al. Reduced risk of colorectal cancer up to 10 years after screening, surveillance or diagnostic colonoscopy. Gastroenterology 2014;146(3):709–17.
10. Kahi CJ, Myers LJ, Slaven JE, et al. Lower endoscopy reduces colorectal cancer incidence in older individuals. Gastroenterology 2014;146(3):718–25.
11. Doubeni CA, Corley DA, Quinn VP, et al. Effectiveness of screening colonoscopy in reducing the risk of death from right and left colon cancer: A large community-based study. Gut 2018;67(2):291–8.
12. Kahi CJ, Pohl H, Myers LJ, et al. Colonoscopy and colorectal cancer mortality in the Veterans Affairs Health Care System: A case-control study. Ann Intern Med 2018;168(7):481–8.
13. Brenner H, Stock C, Hoffmeister M. Effect of screening sigmoidoscopy and screening colonoscopy on colorectal cancer incidence and mortality: Systematic review and meta-analysis of randomised controlled trials and observational studies. BMJ 2014;348:g2467.
14. Weiss NS, Dhillon PK, Etzioni R. Case-control studies of the efficacy of cancer screening: Overcoming bias from nonrandom patterns of screening. Epidemiology 2004;15(4):409–13.
15. Kahi CJ, Imperiale TF, Juliar BE, et al. Effect of screening colonoscopy on colorectal cancer incidence and mortality. Clin Gastroenterol Hepatol 2009;7(7):770–5.
16. Manser CN, Bachmann LM, Brunner J, et al. Colonoscopy screening and carcinoma-related death: A closed cohort study. Gastrointest Endosc 2012;76(1):110–7.
17. Morois S, Cottet V, Racine A, et al. Colonoscopy reduced distal colorectal cancer risk and excess cancer risk associated with family history. Cancer Causes Control 2014;25(10):1329–36.
18. Nishihara R, Wu K, Lochhead P, et al. Long-term colorectal-cancer incidence and mortality after lower endoscopy. N Engl J Med 2013;369(12):1095–105.
19. Garcia-Albéniz X, Hsu J, Bretthauer M, et al. Effectiveness of screening colonoscopy to prevent colorectal cancer among Medicare beneficiaries aged 70 to 79 years: A prospective observational study. Ann Intern Med 2017;166(1):18–26.
20. Brenner H, Altenhofen L, Kretschmann J, et al. Trends in adenoma detection rates during the first 10 years of the German screening colonoscopy program. Gastroenterology 2015;149(2):356–66.e1.
21. Schöttker B, Herder C, Rothenbacher D, et al. Proinflammatory cytokines, adiponectin, and increased risk of primary cardiovascular events in diabetic patients with or without renal dysfunction: Results from the ESTHER study. Diabetes Care 2013;36(6):1703–11.
22. Schöttker B, Haug U, Schomburg L, et al. Strong associations of 25-hydroxyvitamin D levels with all-cause, cardiovascular, cancer and respiratory disease mortality in a large cohort study. Am J Clin Nutr 2013;97(4):782–93.
23. Löw M, Stegmaier C, Ziegler H, et al. [Epidemiological investigations of the chances of preventing, recognizing early and optimally treating chronic diseases in an elderly population (ESTHER study)]. Dtsch Med Wochenschr 2004;129(49):2643–7. German.
24. White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med 2011;30(4):377–99.
25. R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria. (https://www.r-project.org/
). (2019). Accessed October 30, 2020.
26. Brenner H, Hoffmeister M, Arndt V, et al. Protection from right- and left-sided colorectal neoplasms after colonoscopy: Population-based study. J Natl Cancer Inst 2010;102(2):89–95.
27. Thorlacius H, Takeuchi Y, Kanesaka T, et al. Serrated polyps: A concealed but prevalent precursor of colorectal cancer. Scand J Gastroenterol 2017;52(6-7):654–61.
28. Zhao S, Wang S, Pan P, et al. Magnitude, risk factors, and factors associated with adenoma miss rate of tandem colonoscopy: A systematic review and meta-analysis. Gastroenterology 2019;156(6):1661–74.e11.
29. Samadder NJ, Neklason D, Snow A, et al. Clinical and molecular features of post-colonoscopy colorectal cancers. Clin Gastroenterol Hepatol 2019;17:2731–9.e2.
30. Hoffmeister M, Blaker H, Jansen L, et al. Colonoscopy and reduction of colorectal cancer risk by molecular tumor subtypes: A population-based case-control study. Am J Gastroenterol 2020;115(12):2007–16.
31. Allescher HD, Weingart V. Optimizing screening colonoscopy: Strategies and alternatives. Visc Med 2019;35:215–25.
32. Denzer U, Beilenhoff U, Eickhoff A, et al. [S2k guideline: Quality requirements for gastrointestinal endoscopy, AWMF registry no. 021-022]. Z Gastroenterol 2015;53:E1–227. German.
33. Chen C, Läcke E, Stock C, et al. Colonoscopy and sigmoidoscopy use among older adults in different countries: A systematic review. Prev Med 2017;103:33–42.
34. Brenner H, Stock C, Hoffmeister M. In the era of widespread endoscopy use, randomized trials may strongly underestimate the effects of colorectal cancer screening. J Clin Epidemiol 2013;66(10):1144–50.
35. Brenner H, Schrotz-King P, Holleczek B, et al. Declining bowel cancer incidence and mortality in Germany. Dtsch Arztebl Int 2016;113(7):101–6.
36. Siegel RL, Miller KD, Fedewa SA, et al. Colorectal cancer statistics, 2017. CA Cancer J Clin 2017;67(3):177–93.
37. Siegel RL, Torre LA, Soerjomataram I, et al. Global patterns and trends in colorectal cancer incidence in young adults. Gut 2019;68(12):2179–85.
38. Chen C, Stock C, Hoffmeister M, et al. Public health impact of colonoscopy use on colorectal cancer mortality in Germany and the United States. Gastrointest Endosc 2018;87(1):213–21.
39. Guo F, Chen C, Schottker B, et al. Changes in colorectal cancer screening use after introduction of alternative screening offer in Germany: Prospective cohort study. Int J Cancer 2020;146:2423–32.
40. Niedermaier T, Weigl K, Hoffmeister M, et al. Diagnostic performance of flexible sigmoidoscopy combined with fecal immunochemical test in colorectal cancer screening: meta-analysis and modeling. Eur J Epidemiol 2017;32(6):481–93.
41. Ran T, Cheng CY, Misselwitz B, et al. Cost-effectiveness of colorectal cancer screening strategies: A systematic review. Clin Gastroenterol Hepatol 2019;17(10):1969–81.
42. Buskermolen M, Cenin DR, Helsingen LM, et al. Colorectal cancer screening with faecal immunochemical testing, sigmoidoscopy or colonoscopy: A microsimulation modelling study. BMJ 2019;367:15383.