In medical journals, doctors and scientists continue to debate the relative benefits and harms of breast cancer screening for women who are at average risk of developing breast cancer.
This debate is not always reflected in screening programs, most of which strongly recommend mammography to average risk women within a certain age bracket.
Some experts believe that many women are not being provided with complete information about both harms and benefits and cannot therefore make fully informed decisions.
One out of every nine women in Canada will develop breast cancer. This year in Canada, close to 24,000 new cases will be diagnosed and about 5100 women will die from the disease. Doctors can screen women for breast cancer using mammography – a particular type of x-ray image – to try to find cancers in the breast tissue before a lump can be felt by the patient or doctor during a routine breast exam.
In average risk women, who constitute about 90% of new breast cancer cases, the goal of screening is to identify aggressive cancer while it can still be cured. It may seem obvious that breast cancer screening works – after all, many breast cancer survivors have had their cancers detected on routine screening mammography.
But what if most of these cancers would have either done the women no harm, or been detected at a more advanced but curable stage anyway? And what about the women whose mammograms are reported as ‘suspicious’ and then undergo a biopsy, or even surgery and chemotherapy, when these interventions might not have been necessary?
Conflicting evidence, much controversy
A comprehensive review published earlier this year by the Cochrane Collaboration concluded that for every 2000 women invited for breast cancer screening over ten years, just one woman’s life would be prolonged by early detection. At the same time, 10 out of the 2000 women would be ‘overdiagnosed’ and treated unnecessarily with surgery and possibly chemotherapy. More than 200 women would experience psychological distress because of a ‘false positive.’
Not everyone agrees with this interpretation of the data. Another group of researchers analyzed data from two large trials of mammography and concluded that if 1000 women were screened for 20 years, we could expect that 5 or 6 breast cancer deaths would be prevented at a cost of only about 2 cases of unnecessary treatment.
Has policy gotten too far ahead of the evidence?
The benefits and harms of mammography continue to be a subject of debate for researchers and doctors. However, many organizations that issue recommendations for cancer prevention strongly recommend regular screening for average risk women.
For example, Cancer Care Ontario’s website does not mention any possible harms on the main web page, although they do provide information about the benefits and risks of screening on a secondary web page.
Heather Bryant, Vice President, Cancer Control, at the Canadian Partnership Against Cancer says that currently “there is certainly a lot of information out there – the question is whether you are generating more questions than answers in the debate.” Bryant says that “all of the legitimate expert debate may have actually caused more confusion amongst the public” and the “question becomes what is the right amount of information to give to each individual, which will be different for each person. Linda Rabeneck, Vice President of Prevention and Cancer Control at Cancer Care Ontario notes that “part of our job is to make sure that doctors have the information at hand for the shared decision making conversation.”
Incorporating principles of ‘shared decision-making’ for public health initiatives like vaccination and screening may be more challenging than it is for interventions that doctors and patients consider after a disease has already been diagnosed. Within a shared decision making context, the doctor’s role is to interpret and apply the appropriate evidence, considering individual patient’s values, and then together with the patient come to a decision as to whether a particular treatment or test is worthwhile.
Incentives for breast cancer screening
Another reason why women may not end up being fully informed about breast cancer screening is that physicians in Ontario have incentives to increase mammography rates. Family doctors are eligible for bonuses that depend on the proportion of their patients who undergo mammography. These incentives range from $220 for a screening rate of 55% up to $2200 for a screening rate of 75%. In contrast, family doctors who provide complete information to their patients receive no payment if half of their patients choose not to be screened.
Gordon Guyatt, an expert on evidence based medicine and a doctor at McMaster University says “the desirable and undesirable consequences of breast cancer screening for average women are closely enough balanced that it is likely that many fully informed women would decline screening.” He argues that “the government’s current approach ignores this fact and implicitly undermines patient autonomy.”
Is it time to revisit our approach to mammography?
The discordance between the debate in the medical literature and the clear recommendations from some authorities is now the focus of significant attention in the United Kingdom. Mike Richards, the “cancer tsar” for the Department of Health in the UK is commissioning a review of both the evidence as well as how this information is distributed to the public. Richards says this work will “take account of current thinking on how to synthesize information on benefits and harms and how to present these so as to promote informed choice.”
The Canadian Task Force on Preventive Health Care, a federally funded group of experts, is expected to release guidelines later this year.* However, the task force has a relatively low profile, so its recommendations may or may not be viewed as authoritative by all relevant stakeholders.
Cornelia Baines, a doctor and researcher at the Dalla Lana School of Public Health who led one of the mammography trials and has herself had cancer, recently concluded that “it is reasonable for women to choose to be screened, but only if they are completely informed about the probability of benefit versus the probability of harm.” Baines adds however that “people in the field and the media are backing away from facing this controversy.”
* The Canadian Task Force on Preventive Health Care released their recommendations on screening for breast cancer in average-risk women aged 40-74 years on November 22, 2011. You can view the full recommendations here, and related commentary here.
The comments section is closed.
This discussion has been fascinating and I wanted to weigh in with one observation about false positive rates: despite the fact they are seen as anxiety-producing and generally a negative thing for the many women experiencing them, they are valuable for the recruitment and survival of screening programs.
One study of women who were continuously enrolled in the Harvard Pilgrim Health Care plan from July 1983 through June 1995, estimated that 7.4% of them had at least one false-positive at the first mammogram, 26.0% by the fifth mammogram, and 43.1% by the ninth mammogram.
You would think that the packs of wolves crying all over the place would reduce the subsequent rates of screening, but no, it actually increases it. A population-based study from 2003 found that women who had a false-positive mammogram had higher rates of rescreening at 18- and 30-month follow-up. (http://www.jfponline.com/Pages.asp?AID=1508)
“Cancer scares” can be incredibly life-changing events. I’m not a mathematician but the simple logic of this is that high rates of false positive findings, far from being benign, are essential to keep your patrons returning. If you were designing a screening program where you wanted many repeat customers, it wouldn’t make sense to reduce or eliminate false positives from the program.
When I absorb the back-and-forth on this topic I realize that ‘crying wolf’ means many different things to the ears that hear those cries.
If we were to follow the new guidelines, in three years I may or may not be alive. I am 47 years old and have been recently ( Sept 2011) diagnosed with an early, however, relatively aggressive, breast cancer. I was only able to discover my cancer through early screening through mammography testing which I went through almost every year since the age of 40. With any testing, the experience and worry while waiting for results is not always “fun”, however, the benefit of early detection for me, my family, friends, colleagues and employer far outweigh the alternative of a more invasive spread of cancer and potential death. Life is worth living and fighting for so some minor testing annoyance and temporary discomfort from treatment is well worth.
I am diametrically opposed to the findings and recommendations made by the Canadian Medical Association and other pundits that support these new views to officially start mammography testing in women after the age of 50!!!!!
I expect to get through surgery, chemo therapy, hormone therapy and radiation with a significantly lower change of recurrence, spread or death – from 38% down to less than 5% chance of recurrence or spread and less than 5% chance of death – if I make it through the next 5 years, I will have a very high chance of living to bright old age where I will be able see my children grow and continue to be a strong contributing member of society!!
@ John Abbott
Do be careful what you invite the political people in on. Good sense and logic are not their guiding light for any of us. Money is! Their experts are owned and operated by Big Pharma – you know the guys who do not bat an eye at a $3 Billion dollar fine for their misconduct. Mammograms are just another arm of that unpleasant octopus. Being a man you might not know that :o)
http://articles.mercola.com/sites/articles/archive/2011/12/01/glaxosmithkline-to-pay-damages-for-avandia.aspx?e_cid=20111201_DNL_art_1
After the Canadian Task Force on Preventive Health Care released the recommendations, several organizations and experts spoke out against them, saying they could lead to more cancer deaths. The task force is comprised of primary care experts who lead the development of recommendations for clinical preventive services in Canada. – and this story truly highlights the importance of and need for these types of guidelines.
This patchwork of practices and guidelines can lead to confusion for doctors and their patients. Provinces are currently responsible for breast cancer screening programs, and screening practices vary: some provinces (like B.C., Alberta and Nova Scotia) regularly perform mammograms on women in their forties, while others (like Ontario) do not. In the absence of national clinical practice guidelines such as for cancer screening, confusion reigns, patient care is compromised and health resources are either over- or under-used.
In Canada, we have to come together to support a national approach to these and other clinical guidelines; patients and their representatives need to be more involved, and dissemination and take-up encouraged. Recent discussions by our premiers show that this is a national priority calling for action – an action we support based on our work at the Health Council of Canada.
Let’s do this thought experiment: suppose there were no screening programs right now, but we had all of the evidence thus far accumulated. The question is whether to launch a screening program. Would anyone without a direct financial stake in the screening industry say yes on the basis of the existing evidence? I think not. The only reason we continue to have a debate about this is because, as Charles Wright said, screening programs exist, and were launched (at least partly) because of more promising early evidence of the prospect of savng lives while avoiding harm. In health care it is far easier to implement and spread an innovation based on early and hopeful findings than to jettison a program that later research has shown to be less effective and/or more harmful than previously believed. Interestingly, the jig has been up for 20 years, and still society seems not to be able to face the evidentiary music.
And one cannot overlook the gender aspect: breast cancer is (amost exclusively) a women’s health issue, and given health care’s historical bias towards men’s issues, finally an important female condition “had its day”. In the early days one could hardly argue against it, but now a hard look at the evidence leads to the almost inescapable conclusion that all of the mobilization of communities, the runs and the marches, the ribbons and the fundraising have mainly created a huge false consciousness. It’s not the first time a combination of preliminary data, hype, and marketing have shaken loose billions of dollars in search of magic, and it won’t be the last. If all that money had been spent on reducing breast cancer risk factors, far more women’s lives would have been saved. Plus ca change….
Re. Imaging: it is not about the quality of the imaging. The problem is with early detection. The natural history of breast cancer is not sufficiently understood to know how the detected cancerlike changes will behave in the future.
There is evidence that many of them regress, others remain static, others fluctutate below symptomatic level without causing harm, and others take off and threaten life. Nobody can tell from looking at the cells what they are going to do next. Research is being done to look for markers. It may be the case that the answer to that question does not lie in the cells, but outside them, and possibly outside the host: it may be that some in principle unpredictable future event may trigger dormant cells to progress harmfully, an event such as exposure to radiation, or chemicals, or trauma. It is also suggested that, particularly in the 40-49s, removal of a primary screen-detected tumour causes secondaries, thought to occur early in the history of cancerous changes, to flare, resulting in death that would not otherwise have occurred from the cancer. As I understand it in that age group more screened women die of breast cancer than unscreened women.
Early detection is therefore premature detection, and better imaging will make that problem worse not better, leading to greater, not less overdiagnosis.
The judgment whether it is a good idea to detect such changes early and treat so as to pre-empt their future development, possibly by years, and bearing in mind that many are destined never to progress so that this is not pre-emption but overtreatment, is a value judgment and should be made by the individual whose changes they are and no-one else. Some people express the view that they would prefer to have breast cancer treatment they don’t need, than to leave things alone and wait till symptoms appear even though they may never do so. Others regard that treatment as a very high price to pay for something that may never happen. It is essentially prophylactic surgery, and not for women at high risk, but for women with no known risk factors.
Well, my thinking on that is this. The debate about mortality reduction, rates of overdiagnosis and mastectomy, relates to the viability of screening programmes. Viability depends on uptake, uptake depends on whether clients think screening will on balance benefit them, and that depends on what they understand about screening.
As things stand women are not given to understand the nature of the undertaking. In order to be in the running for a chance to prevent early death from breast cancer a woman has to take the risk of being overdiagnosed, and that almost always means being overtreated, because women are not given to understand what overdiagnosis is before they are screened and because it is too late in a brief consultation under pressure of death and by doctors to commit to life changing treatment to start taking that in.
The risk of harm by overdiagnosis is inherent in screening. To secure uptake screeners have not given adequate information about it because they know that if women understood it not enough might attend. Any risk of overdiagnosis will deter some, whatever the chance of preventing breast cancer death; if the risk is as large as some estimates claim, more would be deterred. The success of the programme is threatened by the possibility that those estimates are correct, but that cannot be a reason to reject the estimates, any more than an adverse effect on uptake can be a reason to withhold an explanation of overdiagnosis.
That is the situation in the UK with a NHS screening programme free at point of delivery. In addition to the inherent overdiagnosis, research has failed to definitively demonstrate a reduction in breast cancer mortality, which if it exists, is smaller than is claimed by those who offer screening; and has failed to demonstrate a reduction in all-cause mortality, meaning that even if there are screened women whose death from breast cancer is prevented, they, or some other screened woman not destined to die of breast cancer anyway, has died of something else, and quite possibly the consequences of breast cancer treatment: radiotherapy affects the heart, surgery has its related risks, drug therapies induce emboli and stroke, for example.
So there is the point about informed consent. There is the point about whether a screening programme is worth offering, and that depends on uptake, and on how much overdiagnosis providers tolerate.
Personally, I think there is an ethical question over whether it is ethical to conduct screening which runs on the principle of preventing some people’s deaths (if it did, which, as I have said, is not proven) knowing that other healthy people will receive treatment they don’t need. Informed consent at least allows that only those willing to take the risk of being overdiagnosed for the sake of the chance of preventing death will do so. I am not sure, however, that if I were a practitioner, I could accept that – and that is a value issue, and involves wider issues about how we see the place of medical interventions in human life, whether they are to help the sick, or a technology to give us greater control over our agendas, and how we see ourselves existentially as human beings in relation to each other and to the universe.
The same issues in principle arise over prostate cancer screening, which in the US has overtreated many thousands of men, leaving a huge number incontinent and impotent, a very high price to pay for life, but they have paid it for nothing. That is overdiagnosis. They didn’t get it either, and most of them are walking around thinking “Thank heavens they caught it early or I’d be dead” when that is not true.
And proper guidance cannot be given by doctors being paid for recommendations. It is too open to being a conflict of interest where interest in money overrides interest in patients.
Jeremy: My estimates of over-diagnosis are solely based on peer-reviewed literature. The denominator is the number of screen-detected cancers and the estimated proportion of screen detected cancers that are over-diagnosed goes as high as 50%.
The references are Mandelblatt et al, Ann Int Medicine 2009;151:738-47; Welch et al. J Natl Cancer Inst. 2010;102:605-613; Croswell et al. Semin Oncol 2010;37:202-215; and Jorgensen et al. BMJ 2009;339:b2587.
As indicated by the Canadian Task Force for women age 40-49, screening 2100 women for 11 years will yield 10 instances of overdiagnosis. Obviously too many! And these estimates are congruent with the USPSTF and the Cochrane reviews, both hugely excoriated by screening advocates led by the imaging industry.
Cornelia: I think our initial disagreement boils down to whether we’re talking about over-diagnosis (false positive) or over-treatment (lumpectomy/mastectomy), which many people run together. Over-diagnosis rates are quite high, however over-diagnosis rarely leads to over-treatment. The Canadian Task Force estimates only 10 instances of over-treatment in 2100 women over ten years. That means the rate of over-treatment is less than 0.5%. That is not negligible, but it’s not enormous either. It seems to me that the Canadian Task Force may have taken the right approach in offering different recommendations to women 40-49 than women 50-74, since the prevalence of cancer amongst younger women is so much lower.
However, I think we probably disagree on a more fundamental level. It is not obvious to me, as it clearly is to you, that 10 instances of over-treatment per 2100 women screened is too many to justify recommending screening. My reason for this is simple: for every 10 cases of over-treatment, we would also save a woman’s life. It is not obvious to me that the ethical arithmetic is all that simple here. Mastectomy is a brutal surgery, certainly. But is the harm caused by ten of these surgeries worse that the harm of a woman dieing prematurely? There are, of course, other harms associated with misdiagnosis (fear, distress), just as there are other benefits to screening (earlier detection can result in less severe interventions). All of this makes, for me, a very complicated ethical problem.
Cornelia: Apologies, I use the term over-diagnosis incorrectly in my earlier comment to refer to false positives, glossing over its technical meaning. I’m a non-physician, so I sometimes misunderstand term from the medical field. I do still think it is important to keep in mind how low the frequency of overdiagnosis is in terms of those screened (what I referred to problematically as an overtreatment rate), since I think much of the public thinks in these terms (“how likely is it that I will wind up getting a treatment that I don’t need?”).
Looks like the ‘One size fits all” attitude does not begin and end with MS. It seems to include breast cancer As well. I have had breast cancer, dxed in 1990. I found the lump myself. It was quite visible. The doctor insisted on a mammogram. He knew when he insisted that I had breast implants but he insisted anyway. The line from above, “These incentives range from $220 for a screening rate of 55% up to $2200 for a screening rate of 75%.” pretty much tells me why he insisted,if indeed, incentives were offered in 1990. If they weren’t, well my mind runs off to incompetent so we won’t go there.
If you have implants DO NOT agree to a mammogram. It will show nothing in the way of cancerous growth. Implants show up as black screen. You could be half dead with cancer and a mammogram will give no indication of it at all if you have implants. I know people with implants are among a minority but I also know no one will tell them that little bit of information if myself and two others I know with them are any indication.
I have no idea whether a mammogram can detect cancer or not. I do question the medieval torture tactics being used to possibly find it. And it does not exactly fly by me without my noticing that it is a treatment for women. I can guarantee you men would not race off to get a similar test for testicular cancer.
I was also told I could die if I didn’t follow my various doctors’ guidance. I did up to the operation. I left the hospital after 3 days instead of the 7 they wanted to keep me – university trained nurses do not know how to be nurses anymore. They storm around like bulls in a china shop waking the entire ward just to get a B/P. Then they said chemo and I gave a categoric NO. I did agree to radiation. I did not die from leaving sooner and I did not die from refusing chemo. I gotta tell you, I resent heartily the scare tactics that were used on me and that I gave in to many of them.
Now I have MS and I am running into the same thing – doctors practicing second rate medicine and selling their pride to the pharmaceutical companies with the blessings of our various governments. If you do not see a pattern, try doing your own research. Start with Dr.Burzynski’s cancer treatment. Granted it is for brain cancer but the game they are playing with people’s lives is the same. It is disgusting and it is shameful.
When your government starts crying about financial problems to you, keep in mind that they waste more money on nonsense research and useless trials than on REAL research. It’s useless because regardless of the trials outcomes, they will push those drugs on you anyway – just ask an MSer! Oh yeah, drugs have precedence over surgical procedures. Ask any MSer about that too. Think about all the women who have had almost useless mammograms, not occasionally when they have reason to wonder or worry but yearly, because their doctor will get a cheque if his numbers are high enough. Think about the chemo given to people who in fact do not need it. Go through all the conditions you can think of and count off similar wastages created in the name of that condition. We could be a rich and healthier population if we ran the country instead of the Pharmaceutical companies and those that give unneeded tests using expensive machines – like the mammogram.
The Cochrane review states that 35% more surgeries have been performed in the screened group compared to the non-screened group. The mortality due to breast cancer on the other hand has been reduced by 19%. Somebody could say that this roughly gives 16% more surgeries i.e. overdiagnosis without benefit. However, this review includes 8 quite old clinical trials and the mortality as well as the surgery rates have since changed as better treatment options are available. The best way to answer this question will be to have more recent clinical trials with long follow-up period; however, it is very difficult to conduct a clinical trial now as many countries have already implemented screening programmes or even if not it will be very difficult to recruit patients to the non-screened group and it would raise ethical issues.
Nobody can be happy about the now compelling evidence on the serious problems associated with screening mammography, but its publication is certainly in the best interests of women everywhere. The hope that breast screening could cause a reduction in the mortality from this terrible disease was at first well placed 40 years ago, but it is no longer possible to advocate for an intervention that carries such a tiny likelihood of benefit along with such a huge burden of unintended and harmful consequences in the form of false positives resulting in repeat imaging studies, biopsies, surgery and even the full range of cancer treatment for a histological problem that would never have caused clinical disease. The very essence of science is about seeking truth through the constant cycle of hypothesis, evidence, analysis and revision. In response to a hostile question, John Maynard Keynes famously remarked “When the facts change, I change my mind. What do you do, sir?”
I am acutely aware of the likely reaction to all of this – public, professional and political – but I think that Cancer Care Ontario, in its position of leadership in this area of health care, really must begin to deal with it. As suggested by the Cochrane Collaboration group, the very least response should involve progressive public education including an informed consent form for screening that reflects the reality from evidence rather than the current inappropriate enthusiasm. With this in place in the meantime a comprehensive plan could be developed to deal with the wider implications on the dismantling of population breast screening programs. It would be very unfortunate if the current false hope and vested interests within the screening mammography industry were allowed to prevail.
As I understand it, the estimates of overdiagnosis for the UK range from that given by the NHS at 1 overdiagnosis for 2.5 deaths prevented (though not in their leaflets to women), to that given by the Nordic Cochrane Centre at 1 death prevented for 10 overdiagnoses. The NHS information leaflets neither name nor explain what overdiagnosis is. This is an egregious omission when overdiagnosis is inherent in the nature of screening. Given the huge difference of estimates made by qualified experts on the rates of overdiagnosis, deaths prevented and mastectomy, there is not enough knowledge to conduct a responsible screening programme, and I would suggest that it is not the public who are confused, but the professionals. I would refer them to the Nordic Cochrane Centre reports, and if they don’t accept them, I would ask: do you doubt the validity of the findings for sound reasons, or do you just not like them? If it is the former, then you are in the position of not being able to settle the dispute over risks and benefits, and hence of not being able to offer a sound evidence basis for screening. If it is the latter, then do the decent thing: the NC figures I suggest, put an end to doubt about the value of screening. Meanwhile, screening is here, and the public better be told what it means for them. There is no problem about upsetting people; they should not be patronised by suggestions that they can’t deal with lack of definitive knowledge. It is professionals who are anxious about what to tell people. Tell people what is true: screeners don’t know enough about the effects of screening to be able to guarantee a chance of benefit even for those willing to gamble their bodily integrity for it (for gamble it is). That is the state of the art. It is not about how much overdiagnosis screeners can tolerate – I find it astonishing that health professionals can tolerate any, frankly. But that is irrelevant; what matters is whether individual women will take the risk of harm for the sake of the chance of benefit. You don’t have to guess the answer to that. Tell them and see what happens. I can tell you I know of women who claim to know the issues and say they still want screening. I can tell you I know of women who don’t, and I am one, and we have our rights. I would never have gambled my bodily integrity for whatever chance of benefit, because no-one could guarantee that I would be a winner and not a loser in that lottery. But I was not informed, and I seem to have been a loser. All I can tell you is: don’t do that to people. Tell them what is (not) known. Let them decide. Their body, their risk, on their terms.
While pleased that the important topic of breast screening has been addressed I must also say that I was greatly disappointed at the content. First because the article does not specify that the one-in-niine risk is a lifetime risk. It is not the risk for a woman at at 20 or 30 or 40. Yet many in this age group interpret this statistic as meaning they have a one in nine chance of getting breast cancer tomorrow. Rhetorically it is supremely effective in arousing levels of fear. Secondly I am amazed that Professor Guyatt can claim in 2011 that harms and benefits are in equipoise. Given the many papers published in the last two or three years in high impact journals (which does not include the Journal of Medical Screening), it has become very obvious that declining breast cancer mortality has been observed even in jurisdictions without screening programs and that the decline can be attributed to improved therapy. I wonder how many women would be calm when told that their chance of getting a mastectomy inicreases 20% if they have a screen detected cancer? How many women actually understand that the probablility of over-diagnosis is well over 25%? I personally would not be happy to be treated for a breast cancer unnecessarily. But let the status quo continue! The harms do not exceed the benefits. And their are no issues of conflict of interest.
Cornelia, I agree completely that unnecessary treatment of breast cancer should be avoided. However, I don’t understand your claim that the probability of over-diagnosis is well over 25%. The research cited here indicates the actual probability between 0.2% and 0.5%, suggesting your estimate of >25% is off-base by two orders of magnitude. An over-diagnosis rate of >25% would mean that more than one out of every four women who gets a mammogram would get a positive result. That doesn’t seem at all right. If you have evidence that this is the case, please share it, but I think you’ll find that the actual rate of over-diagnosis is much lower than you have indicated here. None of this takes away from the extremely important issue raised here, I just think we need to avoid sensationalism in favor of a careful examination of the evidence, and thoughtful discussion of how much over-diagnosis we are willing to tolerate to save a life (not a straightforward question at all).
Jerry – an overdiagnosis rate of 25% would not mean that 1 in 4 women who gets a mammogram is overdiagnosed, it means one in four screen-diagnosed cancers is overdiagnosed. In the UK about 0.75% of women who get a mammogram gets a diagnosis; and it would mean 25% of these is overdiagnosed. In absolute numbers for the UK, about 17000 cancers a year are diagnosed through screening, and it would mean that about 4000 of those are overdiagnosed. If the overdiagnosis rate was 25%. Some estimates put it at 50%. Remember also that most of the others are subject to lead bias – they are detected earlier than they would have been if the woman had not been screened, when the evidence suggests that as these are slow-growing good prognosis cancers the treatment would have been just as successful had she waited until symptoms appeared, and so survival is measured from time of diagnosis and looks longer than it really is. The evidence for this is that mortality from breast cancer appears to be the same in areas where screening is not done as in areas where it is: screening is not reducing breast cancer mortality, so women not screened and finding their own cancers live just as long.
The question of how much overdiagnosis those who run screening programmes may tolerate is irrelevant: the question is how much will the women who are risking it? The answer is: explain the risk to them, and let them decide. As I have said, I know of women who claim that they would still want to be screened if the risk of overdiagnosis were colossal. I personally cannot help but wonder if they understand, and if they overestimate their risk of getting breast cancer, given that screening programmes are for the asymptomatic with no known risk factors. But I also know of women who would be deterred from screening by even a small risk of overdiagnosis. Currently the risk is not explained to women who are screened. That is wrong. Women are either suffering appalling distress when they find out too late that they have an unsound diagnosis; or are going round believing what is not true, that their life has been saved because they were screened, when it has not been, and that it has been saved by the treatment they had as a result of screening, when it has not: either because they have been overdiagnosed, or because their cancer would have come to light later and still been just as effectively treated. And the teams treating them know this, but don’t explain it. That is wrong.
Miriam: You are right on the first point – in reading overdiagnosis, I glossed over its technical meaning and conflated it with false positives. I erroneously assumed this is what Cornelia was referring to.
On your second point, I disagree, at least in part. My comment about tolerance was not directed to those who run screening programs, it was directed at those bodies who are responsible for developing recommendations – bodies like the Canadian Task Force on Preventative Health Care. Obviously, patients must be clearly informed of all relevant risks and they must always be free to make decisions about their medical care. Nevertheless, there is a role for professional recommendations (even fully autonomous patients do, after all, ask doctors what they should do from time to time) – and those with the responsibility for developing such recommendations cannot hide behind patient autonomy. They have to face up to the enormously difficult task of weighting the harm of overtreatment against the harm of avoidable mortality. Patients must always be free to make decisions based on their own values, but this does not remove the responsibility for professional organizations to weigh the evidence and offer some guidance for those who want it.