Comments and Response on the USPSTF Recommendation on Screening for Breast Cancer FREE

IN RESPONSE:

We thank the readers for their many comments and welcome the opportunity to clarify our processes and recommendations. Dr. Dean, by focusing on estimates of the number needed to invite for screening, implies that these estimates were the sole determinant of the USPSTF's conclusion that the net benefit of starting mammography screening in women in their 40s is small compared with starting at a later age. In fact, such information was only one of several lines of evidence that we considered. The analysis of Breast Cancer Surveillance Consortium (BCSC) data, information about the risks of radiation and overdiagnosis, and the Cancer Intervention and Surveillance Modeling Network (CISNET) study also contributed. Dr. Braithwaite may be correct that the USPSTF's rationale for when to start screening was left too open to interpretation; however, it is incorrect for anyone to assume that the USPSTF used (or has ever used) NNS as a backdoor means of rationing care.

The USPSTF appreciates Drs. Gail and Shairer's points about risk assessment. To be useful for decision making, risk-prediction instruments must be feasible for use in clinical practice and must accurately discriminate between women who will develop breast cancer and those who will not. One proposed instrument is a risk calculator based on the Gail model and developed by the National Cancer Institute (available at www.cancer.gov/bcrisktool). We note that this calculator requires the input of information that most women do not necessarily know and that one of the stronger risk factors for breast cancer in women aged 40 to 49 years—dense breast tissue—actually presupposes that the woman has had at least 1 mammography performed at an institution that measures and reports breast density, which is not common. This calculator has limited value for helping women in the general population decide when to start mammography screening.

Dr. Begg expresses disbelief concerning the USPSTF's statement that cost is not a factor in its recommendations. He implies that false-positive results and their consequences were used by the USPSTF as cost surrogates or “opportunity costs.” The USPSTF has repeatedly rejected calls to use cost-effectiveness analysis in its recommendations and did not use them for this recommendation. The model used by the USPSTF was not a cost model, and the analysis was not a cost-effectiveness analysis in disguise. The USPSTF used false-positive mammograms in the same way that colonoscopy was used as a counter for screening-associated risk in the decision analysis (1) that supported the recommendation on screening for colorectal cancer. The USPSTF specifically asked Mandelblatt and colleagues (2) to present an analysis comparing the benefits of annual versus biennial screening. The BCSC data show how harms accrue with each screening. Although there is no “equation” displaying the simultaneous consideration of harms and benefits for a particular population, the USPSTF has elsewhere (3) described the judgment required in making its recommendations.

Because 2 large studies (4 - 5), including one of exceptional quality, failed to show a benefit from teaching women BSE techniques, the USPSTF could not recommend that clinicians spend time teaching it. When communicating this, the USPSTF should have more clearly emphasized the word “teach.” Other approaches to public education, including those suggesting that women be “aware of their bodies,” seem to be more promising ways to promote public health.

Dr. Seewaldt raises several important points. Minority women have not been included in most trials assessing the benefits of mammography. Evidence is growing that cancer incidence differs by age in African-American and white women, and that breast cancer in young African-American women has particularly high malignant potential. Much needs to be learned about how to prevent breast cancer mortality in African-American women. However, even if mammography is universally and frequently performed, it is not likely to address the problem of triple-negative tumors in women, regardless of racial or ethnic background. New early-detection technologies, such as serum markers, coupled with enhanced targeted treatments should be pursued to address this problem.

We agree with Dr. Col and colleagues that more work must be done to understand how women and men make decisions when balancing the benefits and harms of medical tests and to create materials to support such informed decision making. Accomplishing these aims will require cooperative efforts by voluntary agencies (such as the American Cancer Society), governmental agencies (such as the National Institutes of Health), and the Agency for Healthcare Research and Quality, in collaboration with behavioral scientists, medical clinician-researchers, and communication experts.

Finally, we believe that Dr. Dean's assertions are incorrect. An RCT based on number needed to invite is the best design to yield an unbiased estimate of the effectiveness (as opposed to efficacy) of a screening program. Estimates of the effect of mammography on breast cancer mortality from observational studies vary widely: 6.5% in the United Kingdom (6), 19.9% in the Netherlands (7), and 25% in Denmark (8). Separating the effect of screening on breast cancer mortality from the effect of improved treatment is methodologically challenging.

The attention attracted by this recommendation has given the USPSTF an opportunity to examine its processes and messages. Although the language of the recommendation was intended for primary care clinicians, we recognize that it was poorly communicated to the broader health care community and public. Despite this, we reaffirm our finding that periodic mammography starting at age 40 to 49 provides small net health benefit compared with starting at age 50. We recommend that physicians and patients discuss the potential harms and benefits when making the individual, personalized decision about when to start screening.

Diana Pettiti, MD, MPH

Ned Calonge, MPH

U.S. Preventive Services Task Force

Rockville, MD 20850