The NordICC colonoscopy trial

NordICC is the first large randomised trial of screening colonoscopy itself, run in healthy middle-aged adults in Poland, Norway, Sweden and the Netherlands. Being invited to one colonoscopy cut bowel cancer cases by about a fifth over 13 years, but the drop in deaths from bowel cancer was small and could be chance, and there was no sign of longer life overall.

Bowel screening is still worth doing from 45, and colonoscopy prevents bowel cancer when people actually go for it. But NordICC does not show that colonoscopy beats a stool test you will actually do, and its effect on dying of bowel cancer remains unproven.

What the trial did

The Nordic-European Initiative on Colorectal Cancer (NordICC) randomly allocated healthy 55–64-year-olds in Poland, Norway, Sweden and the Netherlands: one was invited to a single screening colonoscopy for every two left to usual care. Follow-up data were analysed for 84,585 of them, in Poland, Norway and Sweden.^[1]

Results are reported by invitation, not by who turned up, because that keeps the comparison randomised. That matters here: only 42% of those invited actually had the colonoscopy, which dilutes every result.

Ten years: fewer cancers, no clear drop in deaths

After 10 years, bowel cancer had been diagnosed in 0.98% of the invited group against 1.20% of usual care, an 18% relative reduction. Deaths from bowel cancer were 0.28% against 0.31%, a difference that was not statistically significant. Deaths from any cause were 11.03% in the invited group and 11.04% in usual care: no signal on lifespan at all.^[2]

Thirteen years: the cancer benefit held, the death result didn't firm up

Thirteen-year results, on 84,583 participants, were published in 2026. Bowel cancer cases were 1.46% against 1.80%, a 19% relative reduction, rising to 45% in a model that assumes everyone invited had been screened. Deaths from bowel cancer moved much less, 0.41% against 0.47%, a difference compatible with no effect and still not significant even under that perfect-attendance model.^[3]

Why the death result is weak

Where the benefit sat. It was almost entirely in the lower bowel (the descending colon) and rectum, with essentially no effect on cancers higher up.
Who benefited. The effect was clear in men but not in women.
Too few deaths to detect. The trial was underpowered for its death question by design: bowel cancer killed 0.47% of the unscreened group, against the 0.82% its designers had assumed, so the deaths needed to show a realistic effect were never there.^[4]
Low turnout. With 42% attendance, the invited group's results mix screened and unscreened people.

Low attendance and missing deaths both work against the trial, so it cannot rule out a real benefit on bowel cancer deaths. Nor can it show one.

What it means for choosing a test

NordICC is the reason not to assume colonoscopy is the only "real" screen. A separate Spanish trial found a stool test every two years no worse than one colonoscopy for bowel cancer deaths; more people did the stool test, but among those who started screening, outcomes were similar. That comparison, and which test to choose, is on Cancer screening.

What to do

Start bowel screening at 45 with the test you will actually complete.
If you choose colonoscopy, go when invited. The trial's benefit came from the people who turned up.