Isn't this fundamentally a statistical issue? With a test sensitivity of 99%, meaning a 1% false positive rate, administering it universally to individuals would generate a significant number of false positives. This influx could overwhelm the system, potentially limiting access for those who genuinely require medical attention.
That's what I'm wondering too. Even if 1%, if the follow up is very expensive, the initial test might be cheap but overall cost might be very expensive / prohibitive.
This is a cancer that is typically detected only after it advances far enough that you will be dead in 6 months (not treatment possible). If we could detect it 2 years sooner it would be treatable and most people could life for many more years. Sure the total cost of a positive will be much higher - between whatever tests to verify it isn't a false positive, and all the treatment it will be a lot more money. However by spending that money many people will live a lot longer.
Maybe you don't like any old people, but I often wish I could show my dad my latest project. This test could have saved his life if we had it 15 years ago.
Technically we just "let people die" all the time.
We could spend more money on road safety (or say, reduce the speed limit to 10 mph) but we don't because the costs are too high.
Now, given the relatively common but almost universally fatal nature of pancreatic cancer, any sort of moderately effective screening is probably worth doing, but the argument that if we don't then some people will die isn't very strong.
If you test enough people, you can reach a point where given you have a positive test, there is say a 90% chance it's a false positive. Taking further action can potentially harm you. Given the danger of further testing, and that many time people see no benefit from treatment, its is completely possible on average it can make die more often to get tested.
Screening recommendations are based on statistical arguments that take false positives (plus the risk of the test itself) into account and calculate whether more people would be helped than harmed.
If the test is cheap, you could run it 2 or 3 times. Then the false positive rate would be pretty low and you could proceed with more intense treatments/diagnostics.
That assumes that what causes the false positive is some kind of analytical noise in the test. The bigger concern is biological noise that would persist if you tested the patient again.
It might still be useful to know you have weird protease activity that isn’t cancer derived, but the more of these tests we do, the more likely it is that for every person, there’ll be at least one non-cancer oddity that looks like cancer signal for at least some test.
