It’s a good point. With any kind of screening test, there’s the tradeoff between cost and benefit. Say you spit into a test tube, they run some fancy-dancy new test that costs 100,000 dollars, and if you’ve got a specific variety of Old Movie Disease which will kill you next month if untreated, they can treat it (at the cost of 1 million bucks), and cure you. If you’re 30 years, let’s say that might buy you 50 years of life (if something else doesn’t kill you first).
Trouble is, it’s so rare that only one in a million people get Old Movie Disease.
So to screen a million people, you spend 100 billion dollars, and get 50 more life-years out of the process. Therefore you’re spending 2 billion dollars per increased life-year.
Now, if you’re that one in a million guy (and I’m sure you are ;)), that’s a pretty damn good payoff. For the other 999,999 people, they (or their insurer, or the government) have paid 100 grand and gotten nothing.
That’s a silly example, but it’s the sort of thinking that goes into deciding on recommendations for things like colonoscopy, mammograms, etc. They weigh the cost versus benefit, and the risks. And the tradeoffs between early detection (faster, cheaper treatment and improvement in quality of life) and later detection (longer, more expensive treatment, diminished economic output, etc.).
If that spit-in-a-tube test cost ten dollars, they’d be more likely to do it on everyone even with the remote likelihood of finding a case of OMD. If OMD merely predisposed you to getting painful hangnails versus killing you, they would be less likely to cover even the 10 dollar test. If it hit one in 1,000 people and made them develop severe headaches, that’s increased risk but decreased negative outcome if untreated, so they weigh that.
Things like screening for PKU (on newborns) would seem to fall into the 10 dollars / kill you next month category. While PKU isn’t terribly common, it’s common enough - and the results if untreated devastating enough, and the test is cheap enough, that they do it on all babies.
Some arguments against things like routine screens for various types of cancer are that yeah, if you’re that one in a million, it’s worth it - but the tests themselves may pose danger (radiation, anesthesia etc.), and they may precipitate other interventions (breast biopsy etc.) that themselves carry risks. There are number crunchers out there that try to produce figures that balance that out: screening cost, likelihood of finding something, cost for follow-on tests, economic losses from those (time off work, injuries, etc.), versus benefits (1/1000 helped, lets that 1/1000 work an extra 10 years, etc.).
I assume that with colonoscopy, that 1 extra life-year is an average over all people screened? As in, you screen 1,000 people, 50 are found to have problems, and by treating it each of those 50 are expected to live 20ish years longer than if they hadn’t been screened? If the colonoscopy were done 10 years earlier, maybe only 10 people would have problems, each of those would gain an extra 35 years, so 350 years / 1000 people is .35 years gained. If it were done 10 years later, 110 people would have problems, each would gain 6 years (less gain because things had 10 years more to progress), or net gain of 660 years, or .66 years gain.
Obviously I’m making those figures up - I’m not any kind of statistician - but they seem to illustrate the conundrum.