Debating the value and effectiveness of mammograms

A Canadian study conducted across more than two decades raises doubt that annual mammogram screenings reduce the risk of death among women between the ages of 40 and 59. What do the latest findings mean for women and their health? Judy Woodruff gets different viewpoints from Dr. Carol Lee of the Memorial Sloan Kettering Cancer Center and Dr. H. Gilbert Welch of the Dartmouth Institute.

Read the Full Transcript


    There's new fuel in the long-running debate about the value of mammograms for some women, and whether too many are being overdiagnosed.

    The research, done over more than two decades, found annual screenings didn't reduce the risk of death among women between the ages of 40 and 59. It also found more than 20 percent of breast cancers detected through those mammograms would have been found otherwise and were not life-threatening.

    The study comes amid questions about who should be screened and how frequently. A government panel recommended most women under 50 could skip yearly mammograms. But several professional societies recommend them for women 40 and above.

    We assess these latest findings with Dr. Gilbert Welch of the Dartmouth Institute. He writes about these issues. He's the author of "Overdiagnosed: Making People Sick in the Pursuit of Health." And Dr. Carol Lee of the Memorial Sloan-Kettering Cancer Center, she's on the Breast Imaging Committee of the American College of Radiology, which criticized the study.

    Welcome to you both to the NewsHour.

    Dr. Welch, to you first. In brief, what did this study find?

  • DR. GILBERT WELCH, The Dartmouth Institute:

    Well, Judy, I think there's two things that are really important about this study.

    And first is that it's about screening mammography, not diagnostic mammography. And a diagnostic mammogram is when a woman finds a new breast lump, goes to a doctor, and her doctor orders a mammogram to figure out what the lump is.

    We all agree diagnostic mammography is a useful tool. The second thing to know about the study is that it compared mammography plus a very careful physical exam done by nurses to a group that just got the careful physical exam. So, in other words, it was testing the value of finding things that could not be felt, really small abnormalities.

    And that's its major finding, that there's no point in finding the really small abnormalities that mammography can find. And that's a very important thing for everyone to recognize.


    And, beyond that, it found that there were problems with these screening mammograms?



    Screening, all screening tests come with some potential for benefit, but they also come with some known harms. And the most familiar to people is the problem of false alarms, worrisome findings that have to go through biopsy and multiple tests before they're put at rest, or maybe they are never put at rest. You're just told, you don't have cancer but you're still abnormal.

    But there's also this new problem we're recognizing, which is finding cancers that will ultimately never matter to the patients. And we call that overdiagnosis. It's an unusual idea, where you have cellular abnormalities that meet the pathologic definition of cancer, and yet they never go forward. They never grow. They regress. They disappear.

    And yet we don't know which ones they are, so we end up treating everybody. And so one of the side effects of screening mammography is, it leads more women to be treated for cancer, some of whom didn't need treatment in the first place. And this study, a long-term follow-up with a randomized trial of screening, is the best way to reduce how often that happens.

    And as you said in your open, about one in five invasive breast cancers detected by mammography turn out to be overdiagnosed.


    Well, let's take these one at the time.

    Let me turn to you, Dr. Carol Lee, because, as we said, the American College of Radiology critical of this study. What about the first finding, that screening mammographies overall don't have a benefit?

  • DR. CAROL LEE, Memorial Sloan-Kettering Cancer Center:

    Well, this study wasn't actually a new study. This was just an update of a study that was first reported about nearly 25 years ago that was — that showed no benefit 25 years ago, and so it's not surprising that it showed no benefit on the updated analysis.

    I think what needs to be recognized is that there are a number of other large randomized prospective studies of screening mammography that do indeed show a benefit in terms of reducing deaths from breast cancer among women who get screened — screening mammography.

    So this is just one study that was really an outlier. And there were several criticisms of the study in terms of how it was designed, how it was conducted 25 years ago, and those same problems were not corrected in the re-analysis, obviously. And so I'm not surprised that these results were similar.

    I think it's very important for women to understand that there are a number of other studies that do show decreased deaths among women who get screened.


    Well, let me turn back to Dr. Welch.

    And it's a lot to ask you to address, but, number one, her reference to the fact that there's a problem with the methodology…




    …, and number two, the fact that there are many other studies that don't find these problems.


    Yes, let's deal with those one at a time.

    And there's been a long history by very few people to try to discredit this study, in part because they don't like the result. And the problem has been that they are suggesting that the randomization wasn't good, that somehow sicker patients got in the mammography group. And the long-term follow-up actually proves that is not right.

    In fact, it's very hard to select 80,000 people into two groups purposefully and then get exactly the same death rate in each year in each group. That's actually a pretty good finding that this was a very well-randomized study.


    Well, let me ask Dr. Lee to respond to that point.




    Well, actually, there were an excess of advanced breast cancers in the screening arm of the trial early on, suggesting that more women with preexisting advanced breast cancers got placed into the screening arm of the study, as compared to the control arm.


    Do you want to respond to that, Dr. Welch?


    Well, I think the simplest thing to say is to suggest everyone look at figure two in the article, which is prima facie evidence that you had pretty good randomization.

    And 25 years later, each year, the death rate in the two groups is exactly the same. That's a pretty good argument for randomization.


    And Dr. Welch, then what about the other point that she made, though, that there are many other studies that show there is a benefit?


    There are not — we should be clear, there are not many other studies. Depending on how you count, there are eight or nine randomized trials of screening mammography.

    There's no study like this. And that's why I started to be clear with what the comparison was. The comparison was in the intervention group. It was a mammogram, a screening mammogram, plus this very careful physical exam. No other study did that.

    The control is just the physical exam. So what is being tested here is the question of, what is the value of finding small abnormalities? And that's the important lesson.


    All right, well, let's…


    There's no value to it.


    All right.

    And, Dr. Lee, do you want to respond there that? And then I have a final question for both of you.


    Yes, I do. I do.

    It's been shown that the death rate from breast cancer in this country has declined, whereas it had been rising steadily up until the mid-'80s. It declined with the introduction of regular screening mammography. And since the mid-80's, the death rate in this country from breast cancer has been reduced by about 30 percent.

    Now, some people will argue that that is because of improvements in treatment. But we know, all of us who take care of women with breast cancer know that it is much more likely that treatment will be successful in achieving a cure when the cancers are caught early in their most treatable stage, than as opposed to when they are advanced and have spread, and mammography can achieve that.


    Let me just finally ask both of you, what should women listening to this debate take away from this? How should they think about mammography going forward, Dr. Welch?


    I would say that they should take away there's a lot of professional disagreement. And I think that, in itself, contains some information.

    We don't disagree about the value of treating really high blood pressure. The professional disagreement tells you that this is a really close call. And I'm not suggesting women shouldn't have mammograms. They should just have the choice. It's a genuine choice. It's a close call. It has probably some benefits, but it also has some harms.


    Dr. Lee?


    I think the — the evidence speaks for itself. The decline in breast cancer deaths in this country over the past 30 years has been due in large part to earlier detections by screening mammography.

    And so certainly having a mammogram or not having a mammogram is an individual choice. Nobody is ever forced to have a mammogram. But I think it's important to recognize that, in actual practice, mammography over the years has saved lives.


    Dr. Carol Lee, Dr. Gilbert Welch.

    And I know women will continue to pay close attention to all of this. Thank you.


    Thank you.

Listen to this Segment