Brian,

I have to admit when I listened to the Science Friday interview, it seemed like he glossed over the issue of only 48% accuracy pretty much. What he said was that basically that depends on where you set the criterion used to decide if the results of the test indicate possible cancer. So he claimed you can adjust the criterion so that it would catch all cases of cancer. Only problem (which he kind of mumbled very quickly and was not picked up by Ira Flatow, who normally is good about catching things like that) is it would create a lot of false positives if you did that. I thought "great. Just like PET scans. Expensive tests that produce a lot of false positives and scare us silly." We've already seen on this board that many places won't even do PET scans because there are so many false positives so I fail to say how adjusting a criterion so that it acthes most cases of new cancers but creates a lot of flase positives would really be helpful.

I am encouraged, though, to know that this other one you discuss is in the works.

Nelie