It’s hard to argue with the findings and conclusions of a new paper in JAMA put forth by Drs. Lydia Pace and Nancy Keating, both physicians with public health degrees and appointments at Harvard-affiliated hospitals. The article, published on April 2, has generated a predictable round of headlines along the lines of “Large Study Finds Little Benefit in Mammography.”
You might, while reading or hearing about this news, wonder about the value of yet another study on breast cancer screening. And you might, if you are following this blog, wonder why I remain convinced that mammography – when done right – has the potential to save many women’s lives and, what’s more, to spare even more from the physical, financial and emotional toll of prolonged treatment for advanced-stage disease.
Why I still think that breast cancer screening is a good idea for most middle-aged women (selected, from a longer list):
1. Several valid studies, most notably that from Sweden, have shown a significant survival benefit of breast cancer screening over the long term. These findings, which demonstrated a benefit to women screened in their forties, received little attention in the news.
2. Mammography is not all the same. It’s not a simple, black-and-white or numeric readout. The “result” depends a lot on the radiologist who interprets the images. Some radiologists, by their training and expertise, deliver lower false positive rates and higher true positive (malignant) “pickup” rates. To say that mammography doesn’t work, based on studies over a population, discounts the potential (and likely) benefit of having the procedure done by experts.
3. Pathology methods have improved over the past three decades. Some doctors, including epidemiologists and PCPs, may not be aware of new tools for evaluating tumors that lessen the risk of over-treating early-stage and indolent tumors.
4. Longer survival is not the only benefit of mammography. Late detection involves risks, and costs. “Screening neglect,” as some researchers call it, adds intensity to needed treatment when patients first seek care for advanced disease. This was the focus of a recent paper in the American Journal of Roentgenology that got little press except for the Cleveland Plains Dealer. The investigators in that careful but retrospective analysis found that among women in their forties, breast cancers detected in routine mammograms were significantly smaller than those detected in women who waited until they felt a lump or had symptoms. That finding was no surprise. But what mattered is that the difference in size of invasive breast cancers found – between screened and unscreened women – translated to less chemotherapy for those screened. The point: finding breast cancer early can reduce the need for toxic and costly treatment.
In reading the new JAMA paper, “A Systematic Assessment…” it seems like the authors are giving a well-prepared talk. Essentially it’s a review of reviews on mammography. Yes, it’s that “meta.” They examined the literature on mammography, going back to 1960 – but with an appropriate emphasis on more recent studies, to address 4 (huge, complex) questions: 1) what is the benefit of mammography screening, and how does it vary by patient age and risk?; 2) what are the harms of mammography screening?; 3) what is known about personalizing screening recommendations? 4) how can patients be supported to make more informed decisions about screening?
This is an ambitious set of questions, to say the least. The tables provided, which are for the most part inconclusive, draw heavily on findings that vary in the era of data collected, methods of analysis, and reasonableness of authors’ assumptions, i.e. validity.
But there is no news on mammography here, except that these two thoughtful investigators carefully reviewed the literature. There are no original data in this ambitious analysis, i.e. there is no new information about mammography’s effectiveness, the false positive rate, the harms of screening, overdiagnosis, etc.
Unfortunately the article, at a glance, may add to the growing perception among journalists, primary care physicians who may not read below the paper’s title, and others – including many ordinary women – that mammography’s effectiveness has been, again, disproved. And so if journalists cover this “story,” as they have and will, our collective memory will incorrectly recall another negative finding, which this is not.
The authors’ main conclusions are that decision aides may be helpful, and that developing better ways of screening for breast cancer would be even better than that. I agree.