The New York Times had quite an interesting article today on cancer biopsy laboratory error rates. A few studies have been kicking around for several years giving some sense of the problem and today's article sums up the research nicely. There is nothing in the studies discussed that indicates that the two kinds of error: transposition error (where one sample is mistakenly substituted for another entirely and a person with a clean biopsy may end up with treatment for cancer and a person with cancer may end up with no treatment) and contamination error (where the results may be similarly confused because of the co-mingling of cells from one of more other cancer biopsy samples) are linked to particular kinds of cells being biopsied. There is, interestingly, a difference in error rate between large independent labs and smaller institutionally affiliated ones.
But, despite a lack of any evidence I can find that the problem is prostate cancer biopsy specific, a prostate cancer specific bill has been introduced for the past few years in Congress and once again in May of this year to require Medicare to reimburse for the DNA linked biopsy system for certain prostate cancer biopsies only. Now, I get it that prostate cancer biopsies are reasonably common among the male Medicare population. Yet, there is no reason to believe that the error rate is linked to the kind of tissue involved in the biopsy. The problem may, in fact, be lab procedural.
It takes a kind of cruel calculus to think that a cervical, uterine, or breast biopsy that might lead to either unnecessary cancer treatment or failure in treatment of such cancer is less concerning to the individual involved than a similar experience with prostate cancer. I also have to wonder, if the prostate cancer biopsy bill were to pass, whether it would lead the way to better cancer biopsy lab practices for all or if it might well relieve pressure to address the concerns of all about cancer biopsy laboratory error rates.