Research Analysis · Health & function
A colonoscopy invitation lowered cancer incidence—not mortality at 13 years
In NordICC, inviting adults to one colonoscopy reduced colorectal-cancer incidence at 13 years. Colorectal-cancer mortality was not significantly lower, and participation limits what the invitation result says about completing screening.
By Lois Rune
Published August 24, 2026 · Updated August 24, 2026 · AI-assistance disclosure
Evidence scope
Human cancer incidence and mortality · large randomized invitation trial
Fewer colorectal cancers, without a demonstrated mortality reduction
At 13 years, colorectal cancer was diagnosed in 1.46% of the invited group and 1.80% of the usual-care group. Colorectal-cancer mortality was 0.41% versus 0.47%; the confidence interval included no difference.
Who and what was tested
84,583 adults age 55–64 in Norway, Poland and Sweden were randomized to an invitation for one colonoscopy or no invitation.
What led
The intention-to-treat invitation result: lower incidence, but no demonstrated reduction in colorectal-cancer mortality.
What remains unknown
Whether these results transfer to today’s screening programs, different populations, or people who complete screening after shared decision-making.
The bottom line
NordICC tested an invitation strategy. It did not randomize everyone to completed colonoscopy: 42% participated in the earlier report. Invitation estimates preserve real-world uptake, while estimates among participants require assumptions about who chooses screening.
What the randomized evidence found
At 13 years, colorectal-cancer incidence was 375 of 28,217 people invited to screening (1.46%) and 912 of 56,366 assigned to usual care (1.80%): relative risk 0.81 (95% CI 0.71–0.90).
Colorectal-cancer mortality was 106 of 28,217 (0.41%) versus 236 of 56,366 (0.47%): relative risk 0.88 (95% CI 0.68–1.08). That interval does not establish a mortality reduction.
Benefit and harm travel together
Benefit measured
About 3.4 fewer colorectal-cancer diagnoses per 1,000 people invited over 13 years.
Mortality result
About 0.6 fewer colorectal-cancer deaths per 1,000; the difference was not statistically demonstrated.
Harms reported
The 10-year report recorded 15 major bleeds after polyp removal, with no perforation or screening-related death within 30 days.
Why screening results are easy to misread
Lead-time bias can make survival after diagnosis look longer because diagnosis happens earlier even if death does not. Length bias makes periodic screening more likely to find slower-growing disease. Overdiagnosis means finding disease that would never have caused symptoms or death.
A per-protocol analysis estimated a larger incidence reduction (RR 0.55) and an imprecise mortality estimate (RR 0.70; 95% CI 0.26–1.25). It is secondary and cannot replace the randomized invitation comparison.
At 10 years, all-cause mortality was virtually identical: 11.03% versus 11.04%. The 13-year record has a linked erratum, so this article excludes sex- and location-subgroup claims until the correction text is fully reconciled.
What this means for a screening decision
Current U.S. guidance recommends colorectal-cancer screening for eligible adults using several possible strategies. The best test depends on age, prior testing, health, preferences and access; this article does not select a test for an individual.
Screening can create both benefits and harms. Eligibility, prior testing, family and medical history, competing health risks, and personal preferences belong in a conversation with a qualified clinician.
Sources, funding and editorial method
Funding and conflicts. The 13-year work was funded by the Norwegian Research Council, Nordic Cancer Union, Norwegian Cancer Society and Health Fund of South-East Norway. Authors disclosed research support and consulting, honoraria or equipment relationships involving endoscopy and diagnostics companies including Fujifilm, Olympus, Norgine, Exact Sciences, Steris and Pentax. These disclosures are context; they do not erase the randomized design.
- NordICC 13-year PubMed record and linked erratum
- Published NordICC erratum
- NordICC 10-year report
- NordICC registry record
- USPSTF colorectal-cancer screening recommendation
- Evidence review and prose drafting were separate. The article uses a frozen claim map with claim-by-claim source support.
- Primary population, outcomes, absolute context, limitations, adverse events, funding and conflicts were checked before publication.
- Lois Rune is responsible for the byline and final editorial framing. AI assisted with research organization and drafting under the published editorial disclosure.
- This is educational reporting, not personal medical advice.
More on screening outcomes and tradeoffs
Another randomized screening trial with a cancer-specific mortality benefit
Low-dose CT reduced lung-cancer mortality in selected high-risk smokers
NLST found fewer lung-cancer deaths with three annual low-dose CT screens than with chest radiography in selected current and former heavy smokers. The benefit came with many positive results that were false positives under the trial definition.
Evidence scopeHuman lung-cancer mortality · large randomized screening trial
A screening trial where benefit and excess diagnoses travel together
PSA screening lowered prostate-cancer mortality—and increased diagnoses
After 21 years in Rotterdam’s ERSPC section, repeated PSA-based screening reduced prostate-cancer mortality and metastatic disease in the primary age group, while producing substantially more prostate-cancer diagnoses.
Evidence scopeHuman prostate-cancer mortality and diagnosis · long randomized screening follow-up
Aspirin follow-up separating cancer signals from proven prevention
ASPREE’s cancer signal changed across longer follow-up
The randomized trial found an unexpected cancer-mortality imbalance. Longer follow-up found no difference in overall cancer incidence and no post-trial mortality legacy effect.
Evidence scopeHuman mortality and cancer outcomes · randomized trial plus observational extension
Article history
- : First published after three evidence-review passes and source/correction verification.
- Sources will be rechecked when guidance or the primary records change.