Comment 75 by Martin Torp Dahl :

I am also sorry to say that it is the wrong answer, try it by using 1000 screenings and then divide them up in two groups, the one that gets cancer and the one that does not. Then try to apply the different probabilities for a positive test to these two groups and get the positive results from both groups. You can then divide the true positives by the total number of positive results.

Isn't that what I did? Might be last night's chilli getting to me. EDIT: Oh yeah. Not much point getting the maths right if you then miss a decimal point at the end of it all. Duh. And now I can't even edit the stupidity.

Comment 77 by Martin Torp Dahl :

And by the way, I am not saying that screening is bad, just that interpretation of the results could be better. And for some tests one has to ask if there is a point in screening everybody.

Of course doctors does not start you on chemo right away, but there is now evidence that many people are treated that does not need treatment.

It is important to be aware of the full picture before one decides what to do.

Absolutely.

And there are two different issues which arise. One relates to the person and one relates to the health system generally. Margaret McCartney picks up this point but doesn't fully press it home.

If you test everyone, it costs a lot of money. If you treat based on false positives, it costs more. Those resources cannot now be used for something else, so a public health system needs to be sure that this is a productive use of its resources. As GfA pointed out earlier in the thread, you could make an attempt to do this in QALY and/or micromorts.

But there is also a human consequence, which differs by individual. The invasiveness of procedures which result from a false positive, and indeed the screening itself, are personal matters. Some of us will regard a tube shoved up the colon as a minor inconvenience, others as a gross violation. Even if you don't have a libertarian sentiment in your head, there must come a point where personal choices matter.