The Lancet MAY 04, 2026

Long-term effects of colonoscopy screening on colorectal cancer incidence and mortality: update from the NordICC randomised trial

Michael Bretthauer, Michal F. Kaminski, Magnus Løberg, et al.

Bottom Line

In this large multicenter randomized trial, a single invitation for screening colonoscopy resulted in a significant reduction in colorectal cancer incidence but failed to reach statistical significance for reducing colorectal cancer-specific mortality at 13 years of follow-up.

Key Findings

1. At 13 years, the intention-to-screen analysis showed a colorectal cancer (CRC) incidence of 1.46% in the screening invitation group versus 1.80% in the no-screening group (risk ratio [RR] 0.81; 95% CI, 0.71–0.90).
2. CRC-specific mortality at 13 years was 0.41% in the invitation group compared to 0.47% in the no-screening group (RR 0.88; 95% CI, 0.68–1.08), failing to reach statistical significance.
3. Per-protocol analysis of those who actually underwent colonoscopy demonstrated a greater reduction in CRC incidence (RR 0.55; 95% CI, 0.33–0.81) and mortality (RR 0.70; 95% CI, 0.26–1.25), though the latter remained non-significant.
4. The observed CRC mortality in the control group (0.47%) was substantially lower than the 0.82% originally anticipated during trial design, which limited the statistical power to detect a significant mortality benefit.

Study Design

Design
RCT
Open-Label
Sample
84,583
Patients
Duration
13 yr
Median
Setting
Multicenter, Europe
Population Men and women aged 55–64 years from the general population
Intervention Invitation for a single screening colonoscopy
Comparator No invitation to screening (usual care)
Outcome Colorectal cancer incidence and colorectal cancer-specific mortality

Study Limitations

Low screening uptake (approximately 42% in the invited arm) significantly diluted the observed intention-to-screen effect.
The trial was underpowered for the mortality endpoint due to lower-than-projected background CRC mortality rates in the control population.
The results reflect a once-only screening colonoscopy, rather than a repeated screening program as is common in clinical practice.
Per-protocol analyses are observational in nature and subject to selection bias, limiting their definitive interpretation.

Clinical Significance

The trial confirms that a single screening colonoscopy effectively prevents CRC incidence in the long term, but the modest intention-to-screen mortality benefit and high rates of non-adherence suggest that the impact of a one-time screening strategy in the real-world population may be more limited than previously estimated by observational studies.

Historical Context

For decades, colonoscopy was considered the gold standard for CRC screening, supported primarily by observational and modeling studies that projected large mortality reductions. The NordICC trial represents the first large-scale, multicenter randomized evidence evaluating colonoscopy efficacy, and its publication has shifted the understanding of screening colonoscopy from a universally assumed high-magnitude mortality reducer to a nuanced intervention dependent heavily on participation rates and long-term surveillance.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

How does the 'adenoma-carcinoma sequence' explain why the NordICC trial observed a significant reduction in colorectal cancer (CRC) incidence even though the mortality reduction was not yet statistically significant?

Key Response

The adenoma-carcinoma sequence describes the slow progression of benign polyps to invasive cancer, often taking 10-15 years. Colonoscopy prevents cancer (reducing incidence) by identifying and removing these polyps (polypectomy). Because the trial results were reported at a median of 12-15 years, there may not have been enough time for the prevention of cancer to translate into a statistically significant reduction in deaths, as patients diagnosed with cancer can often survive for many years due to modern treatments.

Resident
Resident

The NordICC trial reported an 'intention-to-treat' (ITT) risk ratio for CRC death of 0.92 (95% CI, 0.69–1.21). How does the participation rate of 42% in the invited group complicate the application of these findings to a patient who is highly motivated to undergo the procedure?

Key Response

The ITT analysis measures the effectiveness of a screening *program* (the invitation), not the procedure itself. With only 42% participation, the ITT effect is heavily diluted by the 58% of individuals who were invited but never screened. For a motivated patient, the 'per-protocol' or 'adjusted' analysis—which suggests a more robust mortality reduction in those who actually underwent colonoscopy—is more clinically relevant for individual decision-making.

Fellow
Fellow

When comparing NordICC to historical sigmoidoscopy trials (like the UKFSST or SCORE), which showed significant mortality benefits, what role does endoscopist performance (e.g., Adenoma Detection Rate) play in interpreting the NordICC results?

Key Response

The NordICC trial included some endoscopists with Adenoma Detection Rates (ADRs) below the currently accepted 25% threshold. Subgroup analyses in NordICC suggested that the benefit of colonoscopy was significantly higher when performed by endoscopists with higher ADRs. This highlights that colonoscopy is an operator-dependent intervention, and its efficacy as a 'gold standard' relies heavily on the quality of the mucosal examination.

Attending
Attending

In light of the NordICC results, how should we modify our shared decision-making dialogue with patients who are hesitant about the invasiveness of colonoscopy compared to non-invasive options like FIT?

Key Response

The NordICC results provide a 'real-world' counterpoint to the high-efficacy estimates often cited from observational studies. Practitioners should use this data to present a balanced view: colonoscopy is excellent for reducing the *risk of getting* cancer (incidence), but the *mortality* benefit at the population level depends on many factors. This may validate a patient's choice for FIT, which has higher adherence rates and a proven mortality benefit through frequent, repeat testing.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

Analyze the impact of 'contamination' in the control arm of the NordICC trial and explain how the availability of opportunistic screening in the participating countries might lead to a type II error regarding the mortality endpoint.

Key Response

Contamination occurs when individuals in the control (no-invitation) arm undergo the intervention outside the trial. If a significant percentage of the control group received opportunistic colonoscopies or other screening (like FIT), the difference in outcomes between the two groups shrinks. This biases the risk ratio toward the null (1.0), making it more difficult to achieve statistical significance for mortality unless the sample size or follow-up duration is substantially increased.

Journal Editor
Journal Editor

If you were the primary editor, how would you address the discrepancy between the study's primary endpoint findings and the 'per-protocol' secondary analyses in the abstract to prevent media oversimplification?

Key Response

Editors must ensure that the primary ITT analysis is the headline to maintain trial integrity, but they must also demand clear reporting of the adjusted (per-protocol) results to provide context. The 'failure' to reach mortality significance in NordICC was widely misreported as 'colonoscopy doesn't work,' when the data actually showed that the *invitation* failed to change outcomes, largely due to low uptake. Editors must force authors to emphasize that screening only works if patients actually show up.

Guideline Committee
Guideline Committee

Should the NordICC trial’s 15-year update prompt a revision of the USPSTF 'Grade A' recommendation for colonoscopy, which currently favors it as a primary screening modality?

Key Response

The USPSTF currently recommends screening (45-75 years) based on modeled life-years gained. While NordICC is the first large RCT for colonoscopy, its low participation and operator-variable quality mean it may not reflect the optimized screening environment of the US. Most committees (like the ACG or Multi-Society Task Force) would likely maintain the recommendation but perhaps elevate the emphasis on 'any screening is better than no screening,' noting that FIT is a highly effective alternative if colonoscopy uptake is a barrier.

Clinical Landscape

Noteworthy Related Trials

2010

UK Flexible Sigmoidoscopy Screening Trial

n = 170,432 · Lancet

Tested

One-time flexible sigmoidoscopy screening

Population

Adults aged 55 to 64 years

Comparator

No screening

Endpoint

Colorectal cancer incidence and mortality

Key result: A single flexible sigmoidoscopy significantly reduced colorectal cancer incidence and mortality in the long term.
2012

PLCO Cancer Screening Trial

n = 154,934 · NEJM

Tested

Flexible sigmoidoscopy screening

Population

Asymptomatic adults aged 55 to 74 years

Comparator

Usual care

Endpoint

Colorectal cancer incidence and mortality

Key result: Flexible sigmoidoscopy reduced both colorectal cancer incidence and mortality compared to usual care.
2017

SCORE Trial

n = 34,171 · JAMA Intern Med

Tested

Once-only flexible sigmoidoscopy

Population

Average-risk individuals aged 55 to 64

Comparator

No screening

Endpoint

Colorectal cancer mortality

Key result: Once-only flexible sigmoidoscopy was associated with a significant reduction in colorectal cancer mortality over a median follow-up of 10.5 years.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis