Letters

Women's attitudes to false positive mammography results

A formerly clueless patient responds

People in the United States may ignore harms of screening

Findings may not apply to United Kingdom

Authors' reply

A formerly clueless patient responds

EDITORI am a patient who received a diagnosis of low grade ductal carcinoma in situ in 1997, on my 43rd birthday, after obtaining a routine screening mammogram showing a cluster of indeterminate microcalcifications. Although I consider myself informed about women's health, I was ambushed by this news. Like the patients in the study by Schwartz et al,¹ I had never heard of ductal carcinoma in situ until it became a terrifying issue that put my life on hold.

Surveying the literature written for patients makes it easy to understand why someone like me could have missed this. I ransacked it, starting with the copy of Our Bodies, Our Selves² that I grabbed from my bookshelf on the day I came home to an ominous message on my answering machine from the radiology clinic. In the 30 pages about breast cancer, the only comment about suspicious mammograms was buried in a sidebar that had apparently been added in a recent revision and had no referring text. I learnt that most books and pamphlets written for patients assume that a woman's entry into the breast cancer culture starts with the discovery of a lump.

Many have long revision histories that predate the widespread use of mammography. Discussions of screen-detected disease are often meagre and carelessly patched in. On the day I received my diagnosis my surgeon dutifully educated me with a pamphlet entitled Breast Lumps.³ It covered the normal breast, benign and malignant lumps, the simple procedure of self examination of breasts, and what happens after the discovery of a lump. Of course, little of this applied to me. I had what was finally described in a small inset on page 11 as an area of abnormality on a mammogram. The rest of the pamphlet contained a list of treatment options ranging from modified radical mastectomy to hormonal therapy. But this information did not help because it did not tell me which treatment was appropriate for my diagnosis.

My discussions with doctors were also an exercise in frustration. I was routinely told, often in the same appointment, that I have cancer and I do not have cancer. Perhaps the subtleties of ductal carcinoma in situ cannot be adequately conveyed in a typical 15 minute consultation, but the cryptic, garbled, and sometimes alarmist information that I got from my doctors was not good enough to make decisions about treatments or to make peace with myself. The only reliable source of information for me was the world wide web, where I located gateways to the medical literature and discovered that the message about ductal carcinoma in situ is far more hopeful and coherent than anything I had read in the literature for patients or heard from my doctors. Given the web's current state of chaos, this is a time consuming enterprise, but an overwhelmed and frightened patient is highly motivated, especially one whose professional research specialty is information retrieval on the web. There is a critical need for better patient education on this subject. The study by Schwartz et al study supports the conclusion that my experience is, unfortunately, a common one.

Carol Jean Godby, senior research scientist. 
OCLC-Online Computer Library Center, Columbus, OH 43201, USA jean-g{at}ix.netcom.com

People in the United States may ignore harms of screening

EDITORI was excited to see the article by Schwartz et al, but I disagree with the authors in their belief that they have shown that support for breast cancer screening does not depend on unrealistic beliefs about the benefits of mammography.¹

They gave the respondents a choice in completing the sentence "All things being equal, if this 60 year old woman got yearly mammograms for the next 10 years, she would have . . ." between the following answers: "A higher or unchanged chance of dying of breast cancer," or: "A lower chance of dying of breast cancer: By one fifth to one tenthBy one thirdBy a halfReduced to zero." I think the question inadvertently gave away too much information about the range of possible correct answers. To this question only the most naive woman would answer that chances were reduced to nil; and no woman answered, "Reduced to zero," whereas 55% answered "Reduced by a half." Since promotions of screening in the United States have at times made it sound like the risks are thereby reduced to almost nil, I suspect that had respondents been given a choice of "reduced by 10%, 20%, up to 90%, 100%," many would have supplied a much higher guess than 50%, maybe as high as 90%, which would indeed point to a vast overestimation of benefit. I don't have evidence and wish the question had been asked differently.

Screening harms, including the risks of undergoing non-beneficial treatment, are a serious matter. The consensus (which carries almost moral force, sometimes arousing indignation if questioned) in the United States that harms of screening (for prostate cancer, breast cancer, or whatever) should be ignored and cannot be substantial, is extremely peculiar (even though I'm a native Texan) and worthy of inquiry; I would like to understand this cultural imperative better.

People really believe that screening and early treatment must be beneficial, I think more as a matter of logic than evidence. I would like to know if the public in other countries views this differently.

Anne Peticolas, senior systems programmer. 
Austin Automation Center (311), Department of Veterans' Affairs, Austin, TX 78772, USA petico{at}io.com

Findings may not apply to United Kingdom

EDITORThe paper by Schwartz et al was all the more interesting because of the availability of comments from peer reviewers.¹ They questioned the relevance of its findings to an international audience. Approaches to breast screening in the United States are different from those in, for example, the United Kingdom, so does this paper add anything useful?

Screening mammography in the United States is mainly recommended annually or biannually from the age of 40 years.¹ A United States website with a link to the American Cancer Society recommends regular screening from age 20 and a clinical breast examination every one to three years for those aged between 20 and 39. Those aged between 40 and 49 years should have a clinical breast examination every year; those aged 50 or older should have a mammogram and clinical breast examination every year.²

The British NHS breast screening programme is for women aged 50 years and over, who are offered screening by mammography every three years until they are 65 years old; they can then continue screening if they self refer. The website of the British Cancer Research Campaign mentions screening only in the context of the national screening programme, but explains how women can be "breast aware" and advises them to see their doctor if they notice any of the changes listed on the website.³ The chief medical officer in England has written to general practitioners and others that there is no evidence to support the efficacy of breast examination by health professionals of the well woman and that palpation of the breast either by medical or by nursing staff should not be included as part of routine health screening for women.⁴

Given the very different approaches to screening, are the results of the paper by Schwartz et al relevant to the United Kingdom? Could the differences be influenced by the way in which health care is funded in these two countries?

Also, can conclusions be drawn for women in general when some groups were excluded from the study? The subjects of the study by Schwartz et al were wealthier and better educated than the general population in the United States, and almost all were white.¹ But general conclusions are made about what education is needed. This point applies to much research that is carried out. One way in which people are marginalised is that most research methods systematically exclude some groups as subjects, and so their needs or knowledge are not known.

Joyce M Carter, consultant in public health medicine. 
joyce.carter{at}Liverpool-ha.nwest.nhs.uk

Samuel Ghebrehewet, specialist registrar in public health medicine. 
Liverpool Health Authority, Liverpool L3 6AL

Authors' reply

EDITORWe appreciate Godby's comments and wish her all the best. Good data about ductal carcinoma in situ are difficult to find, both for doctors and for patients. Since the natural history of ductal carcinoma in situ is not well understood, discussing prognosis or selecting the best options for treatment is extremely difficult. None the less, the available data are reassuring. Mortality from breast cancer for patients diagnosed with ductal carcinoma in situ is low. Moreover, in the 10 years after diagnosis, women with this disease actually had a lower risk of death from any cause than women in the general population.¹

We also appreciate Peticolas's criticism that our response options do not preclude the possibility that women have unrealistic expectations about the mortality benefit of screening. We wish we had used a broader set of responses. Writing good questions is challenging, and we have learned a valuable lesson. Our belief that women have generally realistic beliefs about mammography, however, comes from responses to several questions: our respondents knew that mammography misses some cancers; they accurately estimated the chance of experiencing a false positive result; and they appreciated that other behaviours such as not smoking conferred a much larger health benefit than screening.

Carter and Ghebrehewet are concerned that our results may not apply to poor, minority, or British women. Whether our findings can be generalised is an open question that we were careful to acknowledge. The take home message of our paper stands. The American women most likely to undergo screening appreciate and accept the risk of experiencing a false positive mammogram. On the other hand, they want more education about the possibility of being diagnosed with ductal carcinoma in situ and what such a diagnosis might mean. This last point warrants special emphasis. As new diagnostic methods increase our capacity for early detection, the issues (and questions) raised by ductal carcinoma in situ will become increasingly relevant to many cancers besides breast cancer.²

Steven Woloshin, assistant professor of medicine. 
Lisa Schwartz, assistant professor of medicine. 
H Gilbert Welch, associate professor of medicine. 
Veterans Administration Outcomes Group, Veterans Administration Medical Center, White River Junction, VT 05009, USA

Harold C Sox, Joseph M Huber professor of medicine. 
Dartmouth Medical School, Lebanon, NH 03756, USA