Long-Term Psychosocial Consequences of False-Positive Screening Mammography

At different occasions when we have presented the present study results we have been asked about the interpretation of the (differences in the) mean scores for the various diagnostic groups; notably the groups of women with false-positive screening. In the paper we have made an attempt to illustrate what such differences mean in terms of item answers. While comparison of mean scores is valid inference, especially when the scores fit a Rasch model, other displays may give a more intuitive illustration of the differences in the psychosocial differences between the diagnostic groups. Specifically, we may calculate what fraction of women with false-positive screening mammography results reported more negative psychosocial consequences than women with normal findings; in the absence of a pre-baseline measurement this is the closest one can get to an estimate of the fraction of women that are (negatively) affected by a false-positive screening result. In the table below we have calculated the fraction (in percentage) of women with false-positive screening mammography results that had a higher scale score than the 95%-percentile for women with normal screening results at the five assessment points and for all 12 psychosocial outcomes. We hope this display will further clarify the results of the present study.

Table: The fraction (in percentage) of women with false-positive screening mammography results that had a higher scale score than the 95%-percentile for women with normal screening results. [annfammed.org/site/misc/TRACK/BrodersenTable.doc]

Competing interests: None declared

Thank you to Brown and to Hersch, Jansen & McCaffery for their comments about our paper "Long-term psychosocial consequences of false-positive screening mammography".(1)

In Denmark, there is also lack of evidence-based and balanced information of the benefits and harms of breast cancer screening with mammography.(2) The information leaflet provided to the invited women underplays the harms and exaggerates the benefits of breast cancer screening.(2) Furthermore, by default the invitation to screening mammography contains a prebooked appointment for screening. Every woman is presumed to consent to screening, but may opt out by contacting the clinic. If a woman simply decides not to turn up at the appointment, she will receive further reminders.(3) This nudging and inadequate information about benefits and harms of screening mammography is in contrast to an informed choice whether to participate or not.

Hersch, Jansen & McCaffery raise an important issue about those women who do not accept a screening invitation. In Denmark, there are two categories of such women: 1) those that actively contact the screening clinics and ask to be opted out of the screening programme forever, 2) those that do not show up to the next screening. Previous research has shown that screening participants do not have the same characteristics as those participating in screening. In most cases, the group of people not accepting an invitation to screening are lower educated, have higher rates of co-morbidity, and thereby also lower remaining lifespan.(4-8) We agree with Hersch, Jansen & McCaffery that it would be important to try also to measure the long-term psychosocial consequences among those not accepting the invitation to screening mammography and compare their scores with the scores from screening participants. However, the possibility of introducing a bias in such a comparison would be great due to the different characteristics plus most likely a much lower survey response rate in the group of women not accepting the invitation to mammography screening.

Reference List
(1) Brodersen J, Siersma VD. Long-Term Psychosocial Consequences of False-Positive Screening Mammography. The Annals of Family Medicine 2013 Mar;11(2):106-15.
(2) Gotzsche PC, Hartling OJ, Nielsen M, Jorgensen KJ, Brodersen J. Invitation til brystkraeftscreening: propaganda eller information? [Invitation to screening for breast cancer: Propaganda or information?]. Ugeskrift for Laeger 2009;171(23):1963.
(3) Ploug T, Holm S, Brodersen J. To nudge or not to nudge: cancer screening programmes and the limits of libertarian paternalism. J Epidemiol Community Health 2012 Jul 5;66(12):1193-6.
(4) Burton MV, Warren R, Price D, Earl H. Psychological predictors of attendance at annual breast screening examinations. British Journal of Cancer 1998 Jun;77(11):2014-9.
(5) Christensen B. Characteristics of attenders and non-attenders at health examinations for ischaemic heart disease in general practice. Scandinavian Journal of Primary Health Care 1995 Mar;13(1):26-31.
(6) Jones A, Cronin PA, Bowen M. Comparison of risk factors for coronary heart disease among attenders and non-attenders at a screening programme. British Journal of General Practice 1993 Sep;43(374):375-7.
(7) Lagerlund M, Sparen P, Thurfjell E, Ekbom A, Lambe M. Predictors of non-attendance in a population-based mammography screening programme; socio-demographic factors and aspects of health behaviour. Eur J Cancer Prev 2000;9(1):25-33.
(8) von Euler-Chelpin M, Olsen AH, Njor S, Jensen A, Vejborg I, Schwartz W, et al. Does educational level determine screening participation? Eur J Cancer Prev 2008 Jun;17(3):273-8.

Competing interests: None declared

To comprehensively evaluate the effects of screening, it is crucial to adequately capture the psychosocial impact as well as medical outcomes, so that these may be weighed up together [1]. Brodersen and Siersma's research into the long-term psychosocial consequences of false-positive screening mammography [2] contributes to the mounting evidence about the unintended harms of breast screening.

A key strength of this work is the use of a well-validated condition-specific questionnaire whose psychometric properties have been thoroughly assessed using Item Response Theory and Rasch models [3]. Another strength is the longitudinal design that included not only women who had false-positive screening results but also women with normal results and others diagnosed with breast cancer. The comparisons between these groups over a three-year follow-up period showed that experiencing a false-positive result led to negative psychosocial consequences at a level that was intermediate between women with normal results and those with cancer [2]. While this is an important finding, another relevant group is missing from this study - namely, women who do not accept the screening invitation. Understanding the long-term psychosocial experiences of these women is also important, and future work should try to include them for comparison with those who are screened [1].

The findings presented in this paper reinforce the need for continuing efforts to improve the way women are informed about mammography screening and to understand how they experience and evaluate the consequences of screening, including harms such as false positives and overdiagnosis [4]. The results lend further support for the move towards informed choice in cancer screening programs - an approach already being adopted and implemented in some countries such as Canada [5] and the United Kingdom [6].

References
[1] McCaffery KJ, Barratt AL. Assessing psychosocial / quality of life outcomes in screening: how do we do it better? J Epidemiol Community Health 2004;58(12):968-70.
[2] Brodersen J, Siersma VD. Long-term psychosocial consequences of false- positive screening mammography. Ann Fam Med 2013;11(2):106-15.
[3] Brodersen J, Thorsen H, Kreiner S. Validation of a condition-specific measure for women having an abnormal screening mammography. Value Health 2007;10(4):294-304.
[4] Hersch J, Jansen J, Barratt A, Irwig L, Houssami N, Howard K, Dhillon H, McCaffery K. Women's views on overdiagnosis in breast cancer screening: a qualitative study. BMJ 2013;346:f158.
[5] Canadian Partnership Against Cancer. Quality determinants of breast cancer screening with mammography in Canada. Toronto: Canadian Partnership Against Cancer, 2013.
[6] Informed Choice about Cancer Screening. Approach to developing information about NHS cancer screening programmes. London, 2012.

Competing interests:   None declared

Thank you for this research, we need more critical analysis of women's cancer screening. For too long women have been told they "should" screen with little real or balanced information on the risks and actual benefits. Informed consent is a legal and ethical requirement for all cancer screening, it is the woman's choice, or should be.... I have declined breast screening, an informed decision. Thankfully, more real information is coming out, long overdue. I recently heard Peter Gotzsche from the Nordic Cochrane Institute speak at the Evidence Live Conference in Oxford. His message was clear: breast screening is harmful and should be stopped and it does not save breasts. The Danes have only been screening women in one region and so have the perfect control group...the result: screening has not reduced deaths from breast cancer. The fall in the death rate is about better treatments, not screening. You reduce your risk of a breast cancer diagnosis by one-third if you DON'T screen, about 50% of screen detected cancers are over-diagnosed. Any benefit of screening is almost certainly removed by women who die from heart attacks and lung cancer as a result of over-treatment with chemo and radiotherapy. (See: Research on the effects of over-treatment in breast screening by Professor Michael Baum, BMJ, 2013) The NCI have produced a summary of ALL of the evidence and you'll find it at their website. It's time ALL information was released to women, we dropped the paternalism, put women first and enabled and encouraged them to make informed decisions, and that might include declining breast screening.

Competing interests:   None declared

Thank you to Scaranelo and to Monticciolo & Monsees for their comments about our paper "Long-term psychosocial consequences of false-positive screening mammography".(1)

In Denmark, all women aged 50-69 are invited to biennial breast screening. The invitations are generated via public databases by using the women's unique civil person numbers. Participation in screening mammography is free in Denmark. When women were recruited to the present survey, screening at Copenhagen took place at one public hospital and the recall clinic was placed at another public hospital. In the county of Funen screening took place at two locations: at the University Hospital in Odense and in a mobile screening unit (a bus). The recall clinic in the county of Funen was also placed at the University Hospital in Odense. At all three clinics, screening was conducted by radiographs only and all screening mammograms were read by two radiologists. One to two weeks after screening a letter was posted saying that: either the result was normal and no further action was needed, or that they needed additional examinations and should make an appointment at the recall clinic. In these letters nothing about BI-RADS categories was mentioned, and in the recall letters nothing was mentioned about suspicion of breast cancer. The recall letters were only posted to women on Mondays, Tuesdays and Wednesdays so women could call the recall clinic at the same day or at the following day after they received the recall letter and thereby make an appointment within a few days. This was also why we had to let the women at assessment time point zero complete the COS-BC questionnaire at the two recall clinics because a lot of the time the women showed up at the recall clinic 1 or 2 days after receiving the recall letter, which made it impossible to post the questionnaire to the woman by ordinary mail. For the rest of the assessments points at 1, 6, 18 and 36 months all women in the survey were posted the COS-BC by ordinary mail and completed the questionnaire at home.

In our study we have defined normal results as all women not recalled for additional abnormalities. Although the Danish mammography screening programme does not use BI-RADS categories, we could envision that our normal results would include BI-RADS 1 and 2 categories. All women recalled for additional examinations were classified as having abnormal results. Our abnormal results are most likely similar to a group that includes BI-RADS 0, BI-RADS 3-5 and women recalled due to a "technical recall".

When John Brodersen conducted his PhD-project "Measuring psychosocial consequences of false-positive screening results - breast cancer as an example"(2) he found that in Denmark there were roughly four different diagnostic pathways for women with abnormal screening mammography results to final diagnosis: 1. clinical mammography (anamnesis, breast examination, 3-projection mammogram, ultrasound) 2. clinical mammography plus needle biopsy 3. clinical mammography plus surgery 4. clinical mammography and asked to come back with 3 - 6 months for an addition clinical mammography (also called early recall group) Therefore, focus group interviews were conducted when developing the COS- BC including all four different diagnostic pathways to cover all relevant aspects of psychosocial consequences of false-positive screening mammography.(3) All women in focus group interviews, regardless of the type of diagnostic pathway, read between the lines that they could have cancer. This happened, despite that the word breast cancer was not mentioned in any of the recall letters. None of the focus group participants found the recall to be an innocent event. The general perspective was that this event was equal to a life crisis like a divorce, losing a relative, etc. For example, some of the focus group participants had contacted their lawyer to get their will testament updated. Some participants had considered selling their house, go on early retirement and move closer to their grandchildren. Finally, some participants had even been at the local cemetery and decide were to be buried. None of these thoughts about their own vulnerability and mortality were possible to put in the COS-BC questionnaire because the COS-BC also had to be completed by women with normal screening results.

In the present survey, the assessments points 1, 6, 18 and 36 months are after the final diagnosis: normal, false-positive or breast cancer. Because the time from abnormal screening mammography to final diagnosis varied substantially within the four diagnostic pathways, women with abnormal screening findings had different lapses of time from abnormal screening result to final diagnosis. However, the four follow-up assessments points for all three groups were equally spaced in time after the date of final diagnosis. We are currently analysing our data according to the four different diagnostic pathways and according to the time from abnormal screening result to final diagnosis. We hope to able to publish these results soon.

In our survey we have not collected information about family history of breast cancer, but as we state in our manuscript: "We did not record family and/or friends' history of breast cancer and previous experience with screening mammography. These are possible confounders. However, because the lifetime risk for breast cancer in Denmark is one in ten, all women have relatives or friends with the diagnosis of breast cancer. Furthermore, the higher age, the higher the probability is with: previous experience with breast screening and; having relatives and/or friends with breast cancer. We have adjusted for age as a possible confounder and thereby also indirectly adjusted for previous experience with breast cancer and screening mammography."

For practical, ethical and legal reasons we could not ask women with abnormal screening results to participate in the survey at the same time as the received their recall letter. Most women having a recall letter will in Denmark attend the recall clinic.(4;5) There were two reasons for non-participation in our survey among the women that attended the clinic: either they refused to participate because they were too scared and worried, or because they were not asked caused by sick leave and holiday among the clinic staff.

Monticciolo & Monsees state: "We also note that women with normal results took the baseline survey at home, whereas women with abnormal results took the survey in the clinic. The authors acknowledge this bias, but state it that it would probably lead to less stress in the group with abnormal results. We believe the opposite to be true". We stated in the manuscript that "Some of the women completing the COS-BC at the recall clinic may have "smartened up" their answers to be polite." The reason why we believe some women could have smartened up the answers was because they were very dependent on the staff at the recall clinics at the moment when they completed the COS-BC questionnaire. They had received the recall letter, they had called the recall clinic to make an appointment and now they had shown up at the recall clinic for additional examinations. But before having these additional examinations and before receiving the results of these additional examinations they were asked to complete the COS-BC. After completing the questionnaire women were told to put the COS-BC into an envelope, close it and hand it to the staff. We could well be wrong in this interpretation (as indicated by Monticciolo & Monsees) and the women actually were more stressed at the recall clinic than they would have been at home in the critical time period after receiving the recall letter and until they went to the recall clinic. However, such a difference does not change the conclusion on the long-term psychosocial consequences because all other assessments were done by mail and therefore at home.

Monticciolo & Monsees also reference past research that "...indicates that nearly all those who experienced a false-positive support screening and want to know their status." In the previous mentioned six focus group interviews similar results were found.(3) However, the gratitude that women felt after a false-positive screening mammography result can be interpreted differently than Monticciolo & Monsees do, like the ambivalent reactions John Brodersen and colleagues have revealed about long-term psychosocial consequences of false-positive screening results in breast(3) and lung cancer screening(6) and in abdominal aorta aneurism screening.(7) And as stated in the paper about lung cancer screening: "The COS-BC part II was developed so that each item included response options indicating 'no change' as an anchor relative to two other options of changes in opposing directions. It is well known that people's values and perceptions of life can change as a result of trauma and existential crisis. These changes can be interpreted by the individual as positive, negative or a combination of both. This has been seen in a qualitative study of cancer patients' reactions one or more years after diagnosis.(8) Another study showed that women diagnosed with breast cancer experienced both positive and negative consequences of their disease.(9) Their positive reactions could be seen as a kind of backward rationality, irrationality or ambivalence. These well-known phenomena are for example described in the psychological theory of cognitive dissonance." And by Gram et al in 1990:(10) "In our study it is noteworthy that women willing to pay the highest amount of money to attend another screening are found among those who experienced a positive screening test, but who did not go through diagnostic surgery. It is also notable that a substantial proportion of the study group reported that this experience had a positive impact on their lives. Some of them stated explicitly that they were grateful for this experience, because they found life more precious afterwards. However, it seems unreasonable to put this on the positive side of the balance sheet of a screening, since first the fear, then the relief, are induced by the same screening."

Researchers and health authorities emphasise the need for developing methods that allow balancing of benefits and harms of screening as "there is no common quantitative summary measure, and even qualitatively a balanced and meaningful presentation is difficult to reach".(11) Being able to balance benefits and harms is crucial because of certain special characteristics of cancer screening: the ambiguity of the natural history of cancer, its target population of healthy citizens, and the fact that for some citizens to benefit from early detection of a treatable cancer, other citizens will be harmed with false-positives, overdiagnosis, or complication of diagnosis and treatment. In our present study we have revealed that false-positive screening mammography result in substantial harm. As stated above there are no accepted methods of how the balance the pros and cons of cancer screening including screening mammography. Therefore, it is not up to us or Monticciolo & Monsees to decide whether the benefits outweigh the harms in mammography screening. This must firstly be a political decision based on best available evidence of benefits and harms of screening mammography. If breast cancer screening with mammography is politically decided to be offered to women in a certain age group, it must secondly be up to the individual woman whether she finds that the benefits outweigh the harms. Therefore, we need to provide women with understandable, evidence based information material about all the benefits and all the harms of screening mammography, so women are able to freely make an informed choice of whether to be screened or not.(12;13)

Reference List
(1) Brodersen J, Siersma VD. Long-Term Psychosocial Consequences of False-Positive Screening Mammography. The Annals of Family Medicine 2013 Mar;11(2):106-15.
(2) Brodersen J. Measuring psychosocial consequences of false-positive screening results - breast cancer as an example. Department of General Practice, Institute of Public Health, Faculty of Health Sciences, University of Copenhagen: Manedsskrift for Praktisk Laegegerning, Copenhagen. ISBN: 87-88638-36-7; 2006.
(3) Brodersen J, Thorsen H. Consequences Of Screening in Breast Cancer (COS-BC): development of a questionnaire. Scand J Prim Health Care 2008 Dec 1;26(4):251-6.
(4) Lynge E. Mammography screening for breast cancer in Copenhagen April 1991-March 1997. Mammography Screening Evaluation Group. APMIS 1998;Supplementum. 83:1-44.
(5) Njor SH, Olsen AH, Bellstrom T, Dyreborg U, Bak M, Axelsson C, et al. Mammography screening in the county of Fyn. November 1993-December 1999. APMIS Suppl 2003;(110):1-33.
(6) Brodersen J, Thorsen H, Kreiner S. Consequences Of Screening in Lung Cancer: Development and Dimensionality of a Questionnaire. Value in Health 2010 Aug 18;13(5):601-12.
(7) Hansson A, Brodersen J, Reventlow S, Pettersson M. Opening Pandora's box: The experiences of having an asymptomatic aortic aneurysm under surveillance. Health, Risk & Society 2012 May 8;14(4):341-59.
(8) Jacobsen B, Jorgensen SD, Jorgensen SE. Kraeft og eksistens - om at leve med kraeft. Dansk Psykologisk Forlag; 1998.
(9) Carver CS, Antoni MH. Finding benefit in breast cancer during the year after diagnosis predicts better adjustment 5 to 8 years after diagnosis. Health Psychol 2004 Nov;23(6):595-8.
(10) Gram IT, Lund E, Slenker SE. Quality of life following a false positive mammogram. British Journal of Cancer 1990 Dec;62(6):1018-22.
(11) EUnetHTA. Work Package 4. HTA Core Model for screening technologies - first public draft. 2011. EUnetHTA.
(12) Gotzsche PC, Hartling OJ, Nielsen M, Brodersen J, Jorgensen KJ. Breast screening: the facts - or maybe not. BMJ 2009 Feb 21;338:446-8.
(13) Thornton H, Edwards A, Baum M. Women need better information about routine mammography. BMJ 2003 Jul 12;327(7406):101-3.

Competing interests: None declared

In their recent article, Drs. Brodersen and Siersma evaluated the effects of false positives at mammography [1]. The authors concluded psychosocial harm may result from a false positive result, even at three years. We agree that anxiety over medical testing is real and important, and that as physicians, we should make every effort to minimize patient anxiety. We should not, however, shy away from studies that bring an important benefit to our patients because of fear of a potential downside, without understanding the balance of the pros and cons.

Unfortunately, many important details are not included in the Brodersen/Siersma manuscript, and therefore, the implications of their results cannot be put into proper clinical context. The vast majority of women recalled have the issue resolved simply, with additional mammographic images or ultrasound. Mixing this subgroup of women with other groups likely exaggerates the estimate of anxiety.

For example, we are not informed as to which women and how many with a false positive result had a family history of breast cancer. This is unfortunate, since they might be expected to have elevated concern and this might skew the results. Similarly, it is likely that short interval follow-up was suggested for some women, but we don't know who they were or how many, and we expect that they might have heightened anxiety. What about women who had a biopsy? These women are fewer (about 1-2% of women screened), but presumably these women would have a much higher level of concern compared with those who simply had an additional mammographic image or ultrasound.

We also note that women with normal results took the baseline survey at home, whereas women with abnormal results took the survey in the clinic. The authors acknowledge this bias, but state it that it would probably lead to less stress in the group with abnormal results. We believe the opposite to be true.

Thanks to mammography screening, women in the United States have experienced a 30 percent reduction in mortality from breast cancer. If long-term anxiety regarding testing is an issue, we should address that concern. Past research indicates that nearly all those who experienced a false-positive support screening and want to know their status [2]. We should not discard screening for breast cancer. Doing so would undermine the very real progress made against this terrible disease and result in thousands of unnecessary breast cancer deaths each year [3].

Debra Monticciolo, MD, FACR President, Society of Breast Imaging Professor and Vice-chair of Radiology Texas A&M University Health Sciences, Temple, Texas

Barbara Monsees, MD, FACR Chair, American College of Radiology Commission on Breast Imaging Ronald and Hanna Evens Professor of Women's Health Mallinckrodt Institute of Radiology Washington University Medical Center, St. Louis

References:
1. Brodersen J, Siersma V.D. Long-term psychosocial consequences of false-positive screening mammography. Ann Fam Med 2013;11:106-115.
2. Schwartz, L, Woloshin, S, Fowler, F and Welch, H.G. Enthusiasm for Cancer Screening in the United States. JAMA. 2004;291(1):71-78. doi:10.1001/jama.291.1.71
3. Hendrick, R.E., Helvie, M. United States Preventive Services Task Force Screening Mammography Recommendations: Science Ignored. American Journal of Roentgenology. 2011;196:W112-W116.

Competing interests:   None declared