There's a specific reason researcher's spend so much time deciding what tests are good screens for doctor's to use. For example, it's worth it to do a colonoscopy and mammogram in people over 50. While most people would think- what's the harm of just testing for everything? Why not do body scans every year? Well here's an example why not to!
A 40 something executive woman came to clinic in a panic. When I walked into the room, she was nearly sweating with fear. She handed me a letter from her work and she explained she'd recently taken part in a job related health screening. "They took my blood and blood pressure, then I just got this in the mail" This letter was a form letter: "based on you blood work, there was an abnormality, please see your doctor immediately" Then there was a space to write in the lab that was abnormal. Penciled in was "CA 19-9" no value, just this. The patient had of course gone online to find out what this was- and found out it is elevated in people with pancreatic cancer. There she sat, tremulous, after days of certainty that she must have pancreatic cancer.
I don't get angry very often, but wanted to call her work up and let them have it. This was ludicrous! What asinine lab director would suggest this specific test to screen for? CA 19-9 is not meant as a blood test to do on random people - WHY? because if I took 1000 normal people with an elevated CA 19-9, 998 would NOT have cancer. If I tested it only in people with pancreatic cancer already, then yes a majority would be elevated- but it has only a 1% positive predictive value.
This poor woman! How could I reassure her not to worry? Were my numbers and stats enough? Would my obvious frustration with her work place for doing such a stupid test help? Or will she forever worry now, wonder if at any moment her pancreas will sprout a tumor? Some may say we should still test- for the 1 in 1000 who will have the cancer. But to what harm? If 998 are forever psychologically altered, then is it really worth it?
No comments:
Post a Comment