
A few years ago a headline did the rounds that unsettled a lot of people: the gold-standard bowel cancer test, finally put through a proper randomised trial, had "barely worked". The trial was real — it's called NordICC — and the headline wasn't exactly wrong. It just left out most of the story.
In May this year the same trial reported again, with thirteen years of follow-up. The result repeats a pattern that comes up across medicine, and that's worth understanding once, properly: fewer cases of cancer, but no clear drop in deaths. Those are different things. Reading them well is neither optimism nor cynicism.
What the trial found
NordICC enrolled 84,583 people aged 55 to 64 in Norway, Poland and Sweden (Kaminski et al., The Lancet, 2026). One in three was randomly chosen to receive an invitation to a single screening colonoscopy; the other two-thirds carried on as usual. Note the word: invitation. In the trial's first report, only 42% of those invited actually turned up for the test (Bretthauer et al., New England Journal of Medicine, 2022). The main analysis compares everyone invited — attenders and non-attenders — with everyone not invited. That's the fair way to judge a screening programme, because real programmes also have people who don't come.
Cases. After thirteen years, 1.46% of the invited group had developed bowel cancer, against 1.80% of the others — a 19% lower risk. Among people who actually had the colonoscopy, the reduction was about 45%. Most of the benefit was in the lower bowel and rectum; cancers higher up, on the right side, were not significantly reduced. The effect was clearer in men than in women.
Deaths. Deaths from bowel cancer were 0.41% in the invited group versus 0.47% — a 12% lower rate, with a range of uncertainty running from 32% lower to 8% higher. In the language of statistics: not significant. Among actual attenders the estimate was 30% lower, with an even wider range.
The 2022 report also counted harms: fifteen major bleeds after polyp removal, no bowel perforations, and no deaths linked to the screening within 30 days. Death from any cause was identical in both groups, at about 11%.
The study was funded by Norwegian and Nordic research councils and cancer charities.
Why cases fall but deaths don't — yet
Colonoscopy prevents cancer by removing polyps before they can turn into one. That shows up quickly, as fewer cases. Preventing deaths is a longer, rarer chain of events: the cancer prevented or caught early has to be one that would otherwise have killed, and you have to wait long enough for that to play out.
And "rarer" is the key word. Across 84,583 people, there were 342 bowel cancer deaths in thirteen years. With numbers that small, a trial simply can't tell a 12% reduction from zero — or from 30%. The uncertainty range here is wide enough to hold a worthwhile benefit and a tiny harm at once. "Not significant" means "we can't be sure", not "it doesn't work".
There's a twist the authors flag themselves: deaths from bowel cancer in the unscreened group were 0.47% — far below the 0.82% expected when the trial was designed. The abstract doesn't explain why; better treatment over those years is an obvious candidate. But when the thing you're trying to prevent becomes rarer on its own, any screening test has a smaller target to hit, and needs more people and more years to prove itself.
What's still uncertain
One invitation, one colonoscopy, with fewer than half accepting. A programme with repeat screening and higher uptake would likely do more — and would also carry more of the harms. Thirteen years may still be too short for deaths; the trial is ongoing, with further analyses planned. Right-sided cancers and women did less well in this data, and the paper's summary doesn't say why. None of that is a reason to dismiss the result. It's a reason to keep counting.
What you can do
If you live in England, this debate is partly beside the point: the NHS bowel screening programme doesn't start with a colonoscopy. From 50 to 74, you're sent a home kit every two years that checks a stool sample for hidden blood; colonoscopy comes only if blood is found. The kit is quick, free, and sits unopened in a drawer in a remarkable number of homes.
So the single most useful action is to do the kit when it arrives. And if bowel cancer runs in your family, ask your doctor whether that changes when or how you should be checked.
Fewer cancers is a real result. Whether it becomes fewer deaths is a question still being answered — by people who have been patiently following 84,000 volunteers for more than a decade. That's what honest evidence looks like while it's still being built.
Curiosity first.
— Dr. Brugal
Related on Dr. Brugal
On getting the most from tests and appointments: The one question to ask your GP that changes the appointment and Four blood tests worth asking your GP about if diabetes is in your family. On why good trials can still disagree: Why scientists still can't agree on fasting — and what would settle it.
References
All studies retrieved via PubMed.
- Kaminski MF, Kalager M, Løberg M, et al. Long-term effects of colonoscopy screening on colorectal cancer incidence and mortality: a multicountry, population-based randomised controlled trial. The Lancet. 2026;407(10541):1787-1795. Population-based randomised trial in people: 84,583 adults aged 55-64, 13 years of follow-up. doi:10.1016/S0140-6736(26)00508-8 (PMID 42102826)
- Bretthauer M, Løberg M, Wieszczy P, et al. Effect of colonoscopy screening on risks of colorectal cancer and related death. New England Journal of Medicine. 2022;387(17):1547-1556. Same randomised trial, 84,585 adults, median follow-up 10 years. doi:10.1056/NEJMoa2208375 (PMID 36214590)
Awareness content — not medical advice. Always talk to your doctor or pharmacist before making changes.