« Commons | Main | echinacea »
June 21, 2007
more on mammograms
Following this chain of evidence on mammograms yielded an interesting confirmation of the unwillingness of the medical profession to see clearly past the data cloud. A study in the Journal of the National Cancer Institute dealing with breast cancer screening was accompanied by an editorial by a Professor Baum, MD. He commented that once during an address he gave on mammographic screening, when he suggested that maybe screening does not benefit the premenopausal woman at all, his comments were not well received. He learned "that some topics, particularly breast cancer screening, do not lend themselves to polite and rational scientific debate."
I have no doubt that he is right. But I will persist in attempting.
In their response to Dr. Baum, the authors of the study commented, "We assumed a benefit for mammography among women aged 50-74 years but also assumed that the benefit is small enough that it is important to do screening correctly." A NNS of 838 to prevent one breast cancer death over 14 years is indeed a small benefit.
It is telling that so much time and ink is spent debating the data, when the very point in question, at least the one that should matter, is assumed without proof. Worse, the NNS is available, and the benefit deemed small enough that better efforts should be made. "It is not worth the effort to do this thing, so let's make a better effort to do it right."
| By Robert Maddox | 04:30 PM