Hacker News Reader

Comment by elric | Hacker News Reader

Considering these tests detect blood, and not all polyps (cancerous or otherwise) bleed, there is no way that this is as accurate as a colonoscopy. Maybe if you take into account that more people are willing to use this test than a colonoscpy, it can be as effective because more people get tested. But that's on population level and not on the individual level.

View on HN

I've done both tests.

One thing the colonoscopy came with but the occult blood stool test didn't: a warning that there was a 1 in 1000-10,000 risk of an accidental tear in the colon leading to peritonitis, an immediate surgical emergency (and life-threatening if undetected and untreated).

So yes, the colonoscopy might pick up some rare non-bleeding cancers and polyps ... but it's also going to lead to unpleasant surgical complications in its own right.

(I have no figures for relative risk: but feel it's worth noting.)

Thing is, even bleeding polyps and cancers don't bleed all the time. Ask me how I know.

So, for simple screening purposes I'd say good enough. If there are any other indications a full colonoscopy would be more reliable.

That's how it works in the Scottish NHS. National stool screening program every two years after you turn 50 (you get mailed a sample tube and send it back to the test centre). If it's a positive, you're called in for a colonoscopy. *But* if you show suspicious symptoms outside of the screening program you get a colonoscopy anyway.

(Source: am at very high risk of hereditary colon cancer, had a colonoscopy when my GP learned of the family history, then screening thereafter. All clear so far ...)

Right, it was only a couple of years back we switched from the old test with the little landing pads to the new one that's a tube. I've only had one of the new ones so far.

Also, should have corrected my comment above... it's two year intervals, as you mentioned.

Lang may yer lum reek, neighbour.

Correct me if I'm wrong, but I believe colonoscopy screenings are every 3 years or so (5?). If the stool sample is performed annually you get a lot more data samples/tests.

That would seem to shift some weight to the stool test.

Correct.

As I read it, this article is more about the inadequate US healthsystem than anything else, in particular because it forgets to mention that iFOBT is free and automatically sent to all between 50/55-75 in several European countries. No diet required. Return envelop included. Lab results received by postal mail after a few weeks.

In the US, guidelines say people of average risk should have a colonoscopy every ten years, making your point even more salient. To give you an idea, regularly scheduled colonoscopy (ie every ten years unless higher risk) has a specificity of 86%, while stool tests performed annually are in the 90s (FIT is 96.4%, gFOBT is 92.5%)

Oh yes, you're absolutely right, colonoscopies aren't risk-free, especially if anaesthesia is involved. I wonder if those risks can be reduced (better equipment? more experienced doctors?).

I've had two colonoscopies 5 years apart because in the first procedure they found and removed polyps, which put me on the side 5-year plan (every 5 years). In the last procedure polyps were also found and removed. Since at least one of the polyps removed in both procedures were benign but possibly capable of becoming malignant, the FIT test by itself is probably not sufficient for me as the only cancer screening test.

Yes, that's the problem. I was hoping these results for the FIT meant the uninvasive test was approaching the value of the invasive one, but clearly not close enough for my own comfort.

Reply on news.ycombinator.com