If you have been reading my thoughts for some time, you know that I am not a big advocate of mammograms. I really wish they would do the job they are supposed to do, but they just don’t.
Another study appeared recently, documenting the unpleasant consequences of well-intentioned mammographic screening. It is not the first study highlighting a dilemma that would be much more comfortable to deny. Years ago two large studies, including nearly one half million women, found that mammograms often found breast cancers that spontaneously disappeared, and that breast self exam did not reduce deaths from breast cancer. Much of the medical community and society at large have tried, pretty successfully, to ignore those powerful findings.
The principle author (H. Gilbert Welch) is someone I respect highly, partly because he thinks so clearly and critically about the application of research in clinical practice. Like me, he is trained in Family Medicine in addition to his research bona fides, so he always considers the “bottom line” outcomes. In the past I have recommend one or another of his books because they contain such clear discussions of the vitally important but infinitely confusing matter of using research to make health care decisions. The concepts he explains so well confused just about every medical student, resident and practicing physician I have worked with through my career. It is no surprise then, that his review of mammography data created a white-hot firestorm of controversy. You can read a simple discussion of the study and its implications in an editorial he wrote- http://www.nytimes.com/2012/11/22/opinion/cancer-survivor-or-victim-of-overdiagnosis.html
The essential findings of the study, reviewing the impact of three decades of mammographic screening in America, were:
- Mammograms did not lower the rate of widespread, life-threatening breast cancer
- Improved breast cancer survivor rates are the result of more effective treatment, not screening
- 1/3 of the diagnoses of cancer by mammogram were needless, consequently –
- 1.3 million women have undergone treatment for breast cancers that would never have caused harm
Unneeded treatment causes significant harm. The Hippocratic Oath begins with the admonition, “First, do no harm”. The obvious conclusion is that we need to find better ways to determine which cancers need to be treated and which do not.
The reaction of much of the medical community has been highly emotional. It has not been reasoned and scientific, like we pretend to be. I admit that I have enjoyed the controversy for a few reasons. Welch’s points are valid, but carry frightening implications about our ignorance. Maybe it is because I was already aware of our ignorance in this way (as well as in some many others), that I see mostly good in the controversy. It is another blow against erroneous assumptions. Sometimes it is necessary stir things up. Personally, having had decades of experience as a lightning-rod for narrow-minded criticism, it is nice to see someone else saying the right thing, accepting that it will be unpopular. I dedicated my medical textbook on homeopathy “to those with the courage to ask questions”, both because it is a rare quality in medicine and because it is essential to fundamental progress.
The universal assumption has been that cancers must be treated, or they will kill you. We have so much data, on so many cancers, refuting this facile falsity, that there is no excuse for sticking our heads in the sand. We urgently need to recognize our previous ignorance and move forward to answer the vital question of when treatment for cancer is needed and for whom.