Our own Dr. James Brenner responds to the latest mammography trial out of Canada…
The Canadian Trial, which has been in the news recently, claims to show new evidence that mammography is less effective than the majority of other trials have indicated. It is important to understand that this is not new evidence, but the same evidence that was rejected nearly twenty years ago when investigators exposed serious issues with the way the trial was conducted. While “naysayers” often object to methods in trials that show results that are disagreeable, the problems with the Canadian trial are so severe that this is not an issue of science, but politics, because the trial cost the Canadian government tens of millions of dollars.
The trial, to its credit, was one of the first designed to show an impact or lack thereof of mammography for younger women. While there are many esoteric issues involved with the set of the study, two major problems in its execution invalidate the results, and can be understood my most people.
The first, is that the trial violated the basic principle of screening mammography in assigning people to different groups. Screening mammography is meant to detect small cancers in asymptomatic women before they can harm anyone. Well intended triage personnel who could feel lumps and even enlarged local lymph nodesin women would assign such women to the screening group in order for them to access care more immediately. While it was understandable that such women needed care immediately, assigning them to the “screening” or asymptomatic group should never have been done if the study was a test of screening mammography. When you “stack the deck” with women who have advanced breast cancer into a group that is supposed to have no physical signs or symptoms of breast cancer, it is no surprise that such agroup will not show an improvement over the other group when you look at the number of women who do not do well, especially when there a limited number of cancers that will be found in both groups.
The second major problem was that the in the first part of the study, the mammograms were not conducted in a state-of-the-art fashion. The technologists were not trained, the images were poor, and even the official reference physicist monitoring the trial warned the investigators that the results would be affected. Imagine trying to take a picture of your house through thick fog and trying to identify a small animal in front. You might be able to do it, but if you had no fog, the ability to identify that small animal would be much greater. Again, it is no surprise that poor quality mammography was not able to show a benefit. During the same period of time, in other countries where the mammograms were more state of the art (and not even as good as now), the results, not surprisingly, showed between a 25% and 44% decreased in mortality, when assigning patients to screening groups was conducted properly.
A recent study in the journal Cancer observed that 70% of women who died from breast cancer in their 40’s at major Harvard teaching hospitals were among the 20% who were not undergoing screening mammography. The evidence regarding the life-saving aspects of mammography is based on long term follow up and more data than virtually any other medical intervention that has been studies in the past half century. Sadly, the Canadian trial designers, have revisited old and rejected data, to try to make it “newsworthy” again. The lay press have not questioned these investigators but simply accepted such misinformation.
It is not coincidental that most major medical organizations, including the American Medical Association, the American College of Obstetrics and Gynecology, and the American Cancer Society have looked at ALL the evidence and concluded that asymptomatic women begin screening mammography by age 40 and have it done every year. The recent controversial 2009 recommendations of the United States Preventative Health Task Force are being revisited and restudied, and in the meantime have been excluded from consideration under the Affordable Care Act by specific legislative language.
You can find a copy of the Canadian trial here…