Education And Debate

Women need better information about routine mammography

BMJ 2003; 327 doi: (Published 10 July 2003) Cite this as: BMJ 2003;327:101
  1. Hazel Thornton, honorary visiting fellow (hazelcagct{at},
  2. Adrian Edwards, reader2,
  3. Michael Baum, emeritus professor of surgery3
  1. 1Department of Epidemiology and Public Health, University of Leicester, Leicester LE1 6TP
  2. 2Department of Primary Care, University of Wales Swansea Clinical School, Swansea, SA2 8PP
  3. 3Academic Department of Surgery, University College, London W1P 7PL
  1. Correspondence to: H Thornton, Saionara, Rowhedge, Colchester CO5 7EA

    Scientists continue to argue about the benefits of breast screening, but ultimately decisions about screening should be made by women themselves. To make this decision, however, women need to fully understand both the benefits and the potential harms

    Although mammographic screening is well established in the United Kingdom and elsewhere, its value continues to be debated. Most of the data on mammographic screening come from a group of clinical trials completed over a decade ago. These are subject to constant reworking, reanalyses, and wrangling between the screening zealots and the screening sceptics. New data are unlikely to emerge, as it is improbable that new screening trials, with a control group that is “left to nature” will be conducted.

    Embedded Image

    Decisions about mammographic screening should be based on all the facts


    Most of the controversy surrounding screening, however, seems to miss the point. The question of whether the benefits of screening outweigh the harms is essentially a value judgment. The problem is that, up to now, this judgment has been made by paternalistic agents of the state rather than by women, supported by their general practitioners or others. In order to make an informed choice on screening, women need to be aware of the range of uncertainties for both the benefits (that is, preventing death from breast cancer) and the harms.

    Evidence of effectiveness

    Public health programmes such as the UK NHS breast screening programme aim to reduce mortality from breast cancer by offering routine mammography to healthy individuals. This is a potential public good. But the public receives highly conflicting messages about the effect of screening. Every new published estimate is hotly debated in the media, and the results are usually couched in terms that most people will misunderstand. The latest report of a “44% reduction in breast cancer mortality in women aged 40-69 years”1 was headlined in the Independent, for example, as “Screening halves breast cancer death rate.'” 2 But the text did not mention that mortality from breast cancer also fell in those who were not screened, which was attributed to improvements in treatment. Nor did the article mention the important caveat in the original report (also included in Otto et al's report3) that until formal methods are developed for “partitioning mortality changes by attribution to screening, treatment or other factors” the percentage reduction in mortality remains unknown and incalculable.1

    The public needs to be told about all the outcomes of screening in terms it can understand. This was achieved, for example in Raffle and colleagues' report on cervical screening, which failed to hit the headlines.4

    Earlier claims for the reduction in relative risk of death from breast cancer among women who are screened have ranged from 63% to 6%.5 6 Findings of meta-analyses of randomised trials have also variedfrom 29%7 to the conclusion in a Cochrane review that an effect has not been shown and that breast cancer mortality is a misleading outcome measure.8 Indeed, all cause mortality and surgical and psychological morbidities are equally important outcomes, especially to the affected individuals. Studies often refer to unnecessary treatments arising from overdiagnosis as “biopsies,” but in reality this may encompass segmental excisions, mastectomies, and even radiotherapy (P G⊘tzsche, personal communication9). Comprehensive outcome measures are needed, together with tools that can assess the potential harms and benefits of breast screening. Until tools are developed that are capable of measuring a wide variety of outcomes, we cannot weigh the evidence satisfactorily.

    Evaluating harm

    Harms are often dismissed as a price worth paying for the perceived general good. Individual women may suffer physical, emotional, social, financial, intergenerational,10 or psychological harm.11 The harm may be temporary, lasting around the time of testing and while waiting for results, but can be lifelong. It is important, therefore, that women are able to understand the potential harms and can make an informed choice for which they are prepared to take responsibility. The UK NHS prostate cancer risk management programme attempts this by offering different levels of information and counselling about the pros and cons of the test (

    The way that information is presented is as important as the information itself. Human rights and ethical imperatives demand that individuals should be treated with dignity and enabled to make informed choices based on “substantial understanding.”12 This objective is rarely met. Slaytor et al reviewed 58 pamphlets used in breast screening programmes and found that they all used relative risk information about the benefits in preference to absolute risk reductions. Six pamphlets incorrectly “advised unequivocally that women who have regular screening mammograms every two years halve their chances of dying of breast cancer.”13 Such framing manipulations are highly persuasive in getting people to take tests.14 Evidence shows that use of explicit rather than ambiguous language results in lower consent rates.15 There is also evidence (from all screening programmes, but mainly from mammography evaluations) that when people are offered more detailed information about their personal risks they are less likely to opt for tests.16 Thus, people are more likely to conclude that they were not fully informed when offered relative risk information in isolation and that they were not put in a position to exercise autonomous choices.

    Assessment of benefit

    Few people appreciate that screening contributes to the rise in incidence of breast cancer. Yet invitations to NHS breast screening use the rise in incidence to justify screening.17 In addition, the invitations do not mention that ductal carcinoma in situ accounts for 20% of cancers detected by screening. Ductal carcinoma in situ has an uncertain natural course and those women who have heard of it find it hard to understand, as do many doctors11; it is an early stage of disease that results in a 40% mastectomy rate. The consequences of diagnosis of this enigmatic, little understood disease are serious for women and the health service. Women and their doctors have to make difficult decisions and evaluations of risk, which occupies much time in the clinics, and expensive research is required on management of the disease.

    Women invited for NHS screening are not told the numbers needed to screen to prevent one death or absolute risks of dying of breast cancer. Research by the US Preventive Task Force found that it is necessary to screen 1224 women aged 40-74 years (or 1792 women younger than 50) for 14 years to prevent one death from breast cancer.18 These are pertinent facts for a woman to know when attempting to decide how to manage her risk. The researchers concluded that the age at which the trade-off between benefit and harm becomes acceptable “is a subjective judgment that cannot be answered on scientific grounds.”

    Improving understanding

    We live in an increasingly risk conscious society where screening has been described as “the institutionalisation of risk.” Some may view this as acceptable if there are public health benefits. Others may focus more on the individual's experience of that risk. To meet the needs that go with the experience of risk, we need to demystify the statistics about relative and absolute risk, risks of mortality and morbidity, lifetime risk, 10 year risks, age related risks, etc. Great strides have been made in exploring how risk information can be conveyed simply, in a balanced way, so that it is useful to people's decision making.15 This approach might begin to change the public's perceptions about the risk of dying of breast cancer and the benefits of mammographic screening.11 Gigerenzer mentions five common misunderstandings that could usefully be clarified in leaflets (box).19

    Five common misconceptions about screening19

    • Screening tests are meant for patients with known symptoms

    • Screening reduces the incidence of breast cancer

    • Early detection implies reduced mortality

    • All breast cancers progress

    • Early detection is always a benefit

    Tensions exist between the demands of the screening industry's “pursuit of good uptake” 20 and properly promoting informed choice of patients as required by the GMC guidelines. Although much research has been done, so far there has been negligible improvement in NHS screening leaflets and public misconceptions. Most women who are screened have neither suffered nor been educated about the reality of the uncertainties, harms, and limitations of screening or the consequences of finding pathology of borderline importance.

    Of course, most screening episodes end with the woman feeling reassured. But how many women realise that their risk of developing breast cancer in any one year remains at 2/1000 and that even this degree of reassurance has to be tempered by an interval cancer rate of about half the incidence?

    The focus of research into screening programmes should not be to improve uptake but to develop flexible decision aids to meet women's desires for balanced information. Although some doctors may be concerned about admitting scientific uncertainty, honesty can enhance patients' respect for the profession. A cultural change is necessary to improve the experience of citizens and enable them to take responsibility for their decisions. Some women may say they don't want this responsibility, but it is part of grown up decision making.21

    Summary points

    • Estimates of the effect of mammographic screening on mortality from breast cancer vary widely

    • All cause mortality and morbidity are also important considerations

    • Women offered screening are given limited information on the potential harms

    • Data are presented in terms that are hard to understand

    • Women must be enabled to make true informed choices about screening


    The information inviting women to screening must be improved. It is unacceptable that women taking tests continue to suffer morbidity and regret because they found out the harms of screening from experience. Research into screening needs to encompass the range of outcomes that are important to people being invited for screening. Unless women are able to make true informed choices, funding for the service will continue to be questioned.


    We thank the BMJ editorial team, the reviewers, Peter G⊘tzsche, and Ole Hartling for their comments.


    • Contributors and sources HT had ductal carcinoma in situ diagnosed by screening in 1991. She was invited to join the UK DCIS trial but felt unable to give informed consent. She has subsequently campaigned for consumer involvement in the whole research process and researched the ethical and practical aspects of such involvement. AE has undertaken research on risk communication, shared decision making and informed consent, particularly focusing on screening programmes. His work includes several published systematic reviews and the development of decision aids to assist informed decision making by consumers. MB helped establish the screening programme in 1987-8 in south east England by setting up the screening and assessment unit in Camberwell and King's College Hospital, London. He also directed the Cancer Research Campaign Breast Cancer Trials Group between 1980 and 2002.

    • Competing interests AE and HT: None declared