Life being complex, many simple principles turn out on examination to be not as simple as at first thought. For example, everyone knows, or thinks that he knows, that prevention is better than cure. But is it always? It is often very difficult to say with certainty.
Three articles in a recent British Medical Journal tackle the vexed question of mammography, whose purpose is to detect cancer of the breast early in its development on the assumption that early detection leads to more effective treatment. The advice to women, therefore, is to get themselves scanned regularly.
This seems straightforward and commonsensical, but in fact the question of whether the light of mammography is worth its candle is devilishly complex. For example, if the treatment of breast cancer has improved (and death rates in Britain have almost halved between 1990 and 2010, thanks mainly to improved treatment rather than to early finding), then the number of cases found by mammography in order to save a single life has to increase. This in turn means that old trials – and all trials to determine the long-term effect of mammography have to be old – may no longer be relevant to the present situation. Trials of mammography are, in effect, always trying to hit a moving target.
The main problem that has bedevilled mammography is that of the false positive: the diagnosis of cancer when in fact there is none. For example, it is estimated that approximately 70,000 women in America are falsely diagnosed with cancer annually by means of mammography, that is to say a half of all those who are diagnosed.
False diagnosis does not give rise just to psychological problems such as stress and anxiety; according to one of the authors in the BMJ it results in physical harm and even death. Women who are wrongly diagnosed undergo unnecessary surgery, from which there is always some (if statistically small) danger, both from the surgery itself and the anaesthesia. More importantly, the majority of those who are falsely diagnosed as having cancer will receive radiotherapy, which itself causes, or at least is statistically associated with, an excess of deaths later in life from lung cancer and coronary artery disease. Again according to this author, for each life saved by mammography by detecting cancer early there are between 1 and 3 deaths caused by its other consequences. That is why trials of mammography that report only on death rates from breast cancer are insufficient and even misleading. As treatment for breast cancer improves, the presumption against mammography only gets stronger: unless the mammography itself improves in accuracy.
The situation is further complicated by the fact that not everyone agrees with these statistics. In the very same edition of the journal, another expert comes to the conclusion that breast screening is worthwhile, despite overdiagnosis. The harms of radiotherapy, for example, are decreasing as techniques are refined; moreover, diagnosis is becoming ever more accurate and sophisticated, allowing treatment to be tailored to the different kinds of cancer from which women suffer.
A third paper points out the ethical dilemmas of doctors in advising their patients. What should they tell them about mammography? If experts who have devoted most of their professional lives to studying the problem cannot agree even on the facts, for example about how great are the harms caused by overdiagnosis, what hope is their for the ordinary doctor who has only lightly touched the subject?
The chances are, for various reasons, that he will advise screening, for it is a fundamental truth that there is more rejoicing by malpractice lawyers over one false negative than over ninety-nine false positives.
First published in PJMedia.