Hi Mickey 191,

there's a fairly readable chapter about mammography etc. in G. Gigerenzer's Calculated risks: How to know when numbers deceive you.

A very sobering account is from M. Roberts (shortly before her dying from breast cancer), Breast screening: Time for a rethink, in: Brit. Medical Journal, 1989, 299, pp1153-1155.

Even if the false positive rate (I have no number at hand) is fairly low (per single woman), in a large scale screening (in which the base rate of true positives is low compared to negatives) the total number of false positive can be irritatingly high.

Hamm et al., Prophylactic mastectomy in women with a high risk of breast cancer, New England Journal of Medicine, 1999, 340, pp 1837-1838, give fairly high values for relative risk reduction for mastectomy. But still, if you calculate their relative numbers into absolute numbers you get that roughly 40 women must undergo mastectomy for 1 to be saved by this procedure. 1 in 40 only! (the others either died despite the procedure, a small number, or, the larger number, wouldn't have died even without mastectomy; they could tell that for they had a control group not undergoing mastectomy despite being in a risk group)

As a kind of weak consolation, your risk was estimated at that time and your then estimated risk has been reduced by mastectomy, though not as much as you would have liked. With more information you might have made another decision then. I'm not against mastectomy, but I'm for better patient information.

Just a story about quite different ways of informing a patient: My wife and I went to a childbirth clinic information evening. When asked, can it happen that your clinic is full when I come, one doctor said you shouldn't worry about that that normally does never happen. The other doctor said, I'm here since three years, we have about 500 births a year and that has happened once in these three years. "Normally never" or "1 in 1500", you may guess by which doctor I felt better informed. That way it went on all evening. One doctor tried to be reassuring with fuzzy words without exact meaning, the other answered whenever possible (and apologised if he had not the information) with hard numbers in an understandable way this way allowing us to make our own evaluation.

Wolfgang