The 1.1% number that’s supposed to end colonoscopies at 75
A study meant to stop surveillance colonoscopies in older adults is getting read as proof it should continue.
By The Weekend · · 4 min read
Among adults 75 and older who'd already had a polyp removed, the 10-year risk of getting colorectal cancer was 1.1%. The risk of dying from it was 0.5%. Meanwhile, roughly half the cohort died of something else entirely within that decade. Three letters to JAMA this month argue about what that gap means — and whether it means the same thing everyone assumed.
1.1% is the number. The letters can't agree what it's for.
The original study, published in JAMA, followed older adults who'd had a colonoscopy and tracked what happened over the next ten years. Among those with a prior adenoma — the precancerous polyp that's the whole reason surveillance colonoscopies exist — colorectal cancer showed up in 1.1% of people. It killed 0.5%. Everything else, from heart disease to the ordinary attrition of being very old, claimed close to half the group.
Three separate letters responded, each pulling a different thread from the same table. One group emphasized that the excess risk tied to having had an adenoma was, in absolute terms, small — smaller than the framing around "high-risk" surveillance implies. A second group agreed the difference was modest but noted it was still measurably higher than in people with no prior adenoma at all. A third group went further, arguing the real finding isn't about adenomas at all — it's that noncolorectal death dwarfs colorectal cancer death at this age, full stop, adenoma or not.
All three read as consensus on the surface. All three want a different clinical response.
The percentage looks small because the denominator is doing work
A 1.1% ten-year incidence sounds negligible, and set against nothing, it is. But surveillance colonoscopy isn't compared against nothing — it's compared against the discomfort, cost, sedation risk, and bowel-prep burden of repeating an invasive procedure on someone who is, on average, closer to the end of their life than the middle of it. The 0.5% death figure is the number that actually settles arguments about whether that trade is worth making, and it's the number all three letters keep circling without landing in the same place.
Here's why the figure reads smaller than it should to some clinicians and larger than it should to others: competing risk. If you're going to die of something else with roughly 50% probability inside ten years, a 0.5% chance of dying from colorectal cancer isn't just small — it's small relative to a denominator that's already been cut in half by everything else that can kill you at 75-plus. That's the argument for stopping surveillance. But the counterargument, made plainly in one of the letters, is that "small relative to competing risk" isn't the same as "zero," and a test that's cheap, familiar, and already scheduled doesn't need to clear a high bar to keep being offered.
The strongest pushback isn't about the number — it's about what replaces it
The most serious objection in the letters doesn't quibble with the 1.1%/0.5% figures at all. It asks what a clinician does instead. Stopping colonoscopy-based surveillance in this population only works as policy if there's something to put in its place, and the study's own authors, replying to the letters, point to noninvasive screening — stool-based tests, primarily — as the substitute already sitting in their original discussion section.
That's a real option, but it's not a free one. Noninvasive tests catch fewer early lesions than colonoscopy does, and they generate false positives that lead back to the same invasive procedure they were meant to avoid. Swapping one test for another doesn't eliminate the burden on an 80-year-old patient — it just changes where in the pathway that burden lands. The authors' reply doesn't dispute this; it treats deimplementation as the goal and noninvasive testing as a reasonable, not risk-free, bridge.
What nobody in the exchange contests is the underlying asymmetry: at this age, the thing most likely to kill you almost certainly isn't the thing the colonoscopy is looking for. That's not a call to abandon screening. It's a call to stop pretending the screening decision is really about cancer risk in isolation, when it's actually a bet about which of several unrelated risks is worth spending a person's remaining years managing.
What the 1.1% does next
None of the four documents — the original study and the three replies — proposes a new guideline. That's deliberate; letters to the editor don't carry that weight, and the authors' reply explicitly frames noninvasive screening as a "possibility," not a directive. But the debate maps onto a real and unresolved gap in US screening guidance: none of the major bodies currently issue an upper age cutoff backed by data this specific, and clinicians are left extrapolating from all-comers screening trials that mostly stopped enrolling people this old.
The next test of this number won't be another letter. It'll be whether the professional societies that write colonoscopy surveillance intervals — typically three, five, or ten years depending on what a prior colonoscopy found — start citing competing-risk data like this one directly in their recommendations, rather than leaving the judgment call to individual gastroenterologists deciding, procedure by procedure, whether an 82-year-old patient's remaining life expectancy still justifies the prep, the sedation, and the follow-up visit for a risk that's already this small next to everything else waiting to happen to them.