Back to Menu

THE ATTACK ON MAMMOGRAPHY - SHOULD IT BE TAKEN SERIOUSLY?

(January 2002; Revised August 2002)

The attack is ferocious and it has generated a huge amount of newspaper and public attention. It all started in January 2000 when two investigators from the Nordic Cochrane Centre in Copenhagen published an article in the prestigious journal The Lancet titled "Is screening for breast cancer with mammography justifiable". It ended "we conclude that screening for breast cancer with mammography is unjustified".

That carries a lot of weight because the Cochrane Group is widely considered the best in providing authoritative reviews on important medical topics. The authors examined eight randomized trials of mammography. They said only two (one in Malmo, Sweden, one in Canada) were acceptable and neither of these showed any reduction in deaths from breast cancer. All the other studies the reviewers rejected on rather narrow grounds are considered solid studies by most experts and showed a clear benefit.

The article provoked a veritable firestorm of angry letters to The Lancet. This, in turn, resulted in a second article by the two authors in The Lancet of October 20, 2001 in which they defended and expanded their position.

Here is what you need to know in evaluating the anti-mammogram attack:

- the methodology and the conclusions of the review (called a meta-analysis) of trials of mammogram screening has been questioned (even repudiated) by experts on breast cancer at the Cochrane Group itself. So, colleagues of these authors in the Cochrane Group seem to have major concerns about the validity of the review.

- This meta-analysis is not the only one that has reviewed the eight randomized mammography trials. One published in 1993 and one published in 1995 both found overall significant reduction in deaths from breast cancer among women undergoing mammography.

- The Canadian study these authors like has been heavily criticized ever since its publication. There are multiple potential flaws, including the finding that physical examination was about as effective in picking up breast cancers as mammograms. That makes no sense at all; mammograms should detect a lot more cancers than physical examination. That strongly suggests that the criticisms of others are correct, namely that the mammogram techniques in the Canadian study were inadequate.

- The screening studies reviewed were all done by intention to treat. That means that, if women were randomized to receive mammograms annually, but never came for the tests, then developed breast cancer and died of it, they were considered mammogram failures, even though they never actually had any mammograms. The idea is that intention to treat better reflects the "real" world in which some women invited to screening will not comply. But, if the goal is to determine whether the mammogram is a good screening technique, it is essential to specifically determine the benefit in women who actually received mammograms.

- The authors used 1988 results from Malmo. No reduction in deaths from breast cancer were found, but the data suggested that benefit would be found with longer followup and a significant reduction in deaths from breast cancer was indeed reported in two followup studies. So, one of the two studies used to prove mammography was ineffective actually showed exactly the opposite. That is an unforgivable and deliberate error because the authors acknowledged they knew about the followup studies.

- The editorial by the editor of The Lancet is extraordinary. It is six paragraphs long. The first five paragraphs are an attack on those in the Cochrane Group who had the temerity to disagree with the attack on mammography published in his journal. There is nothing in the editorial to suggest he has reviewed the science; he is just angry at the Cochrane Group. In the last paragraph, he finally gets around to the issue of mammography and concludes "at present, there is no reliable evidence from large randomized trials to support screening mammography programs". It is a very strange editorial.

Even though those were intention to treat studies, overall, the trials show about a 25 percent reduced risk of dying from breast cancer; if only those who actually had mammograms are considered, the reduction in risk of death from breast cancer increases to 33 percent. Today, the mammography machines and techniques are a lot better, so the benefit is likely to be even greater.

So, what is the bottom line.

1. Ignore the current attack on mammograms. It is based on a flawed and incorrect review.

2. Basically, nothing has changed. There is plenty of evidence mammograms do detect cancers early and reduce the likelihood of death from breast cancer.

3. Every woman starting at age 40 should have an annual mammogram. In future years, we may have new and better screening tests, but for now, the mammogram is the best technique for early detection of breast cancer when the cure rate is very high.

Additionally, women should have annual breast examinations by a health care provider starting at age 40. Healthful Life also continues to recommend monthly self examination of the breasts starting at age 30 (see current editorial on breast self examination).

A new analysis of breast cancer deaths in previously reported Swedish studies plus editorial comment appears in the journal Cancer August 1, 2002. It strongly supports the Healthful Life conclusions. Of those women invited to participate in mammography who actually underwent mammography screenings, deaths from breast cancer were reduced by 44 percent. The longer the women participated in the screenings, the greater the benefit. And, as noted, the benefit is probably even greater today because

- mammogram machines and techniques are better

- the recommended screening is done annually after age 40. In the Swedish and other studies, the interval between screenings was often one and one-half to three years.

Additionally, once the diagnosis is made by mammography or clinical breast examination, our treatment is now considerably more effective than it was a decade ago.

The message is clear - every woman should have an annual mammogram starting at age 40