All of you are by now aware that a recent publication from Canada debunks the value of screening mammograms. The Canadian study is deeply flawed. In 1980 Canadian physicians began to randomize almost 90,000 women ages 40-59 either to annual mammography or not for five years (all women over 50 got careful breast exams as well). After enrollment was complete and each woman in the half who underwent mammography had finished the five procedures, they were followed without further intervention either until death or for up to twenty-five years, ending in 2005. The number of women dying of breast cancer was roughly the same whether or not they had received the five mammograms. From that result has arisen the suggestion that mammography has been oversold to the American woman. Why is the Canadian study flawed? First, of the 1005 women in the study who died from breast cancer, only 351 died if their cancer was diagnosed when they were actually in the five years of active study. The others all developed cancer and died of it after their five years on study. Of the women who got cancer while in the study, 70.6% of the women in the mammography group survived their cancer, while only 62.8% of the non-mammo group did. This difference was highly significant and is typical of what most specialists currently believe is the magnitude of improvement in outcome with the addition of mammography. Furthermore the cancers discovered in the mammo group were smaller than those in the controls. That equal numbers of women in each group died of breast cancer by year 25 is a result of the dilution of these data by the subsequent cancers that developed after the study closed. Interestingly when cancer was discovered in the mammo group 68.2% of these could be felt by the examining physician. This is a much higher number than is seen in contemporary US mammogram centers, where most of the cancers picked up on annual screening cannot be felt. This suggests that the equipment used in the study (the last study mammogram was done in 1988) and/or the skill of the radiologists was inferior to those seen in contemporary US mammogram centers. So even though the diagnostic acumen of mammography was less than currently possible, women whose cancer was diagnosed as a result of mammography survived in greater numbers.
This study is probably important because of the size of the population studied but the design of the trial and the likelihood that the screening techniques used were inferior make it imperative that we view these data with extreme caution. The suggestion by some that mammography be abandoned based on these data is unwarranted and dangerous.
Dr. Stark finishes his comments….So when friends and colleagues ask me what to make of this study, I urge caution and remind them not to abandon forty years of data, which show the value of annual mammography in saving lives