In my last post I introduced some of the controversies surrounding breast (and prostate) cancer screening methods. I’ve been digging into the research on screening mammography for an assignment for the radiology elective I just finished, and realized there is definitely more on this subject that I want to write about.
I’ve been focusing my reading on the perceptions (and misconceptions) about mammography, both on the side of physicians and patients (though breast cancer awareness has become such a public issue, I wish there was research looking at general awareness about cancer, not just awareness in women of screening age- but I digress…).
So how effective is mammography?
Over the years, quite a lot of data has been generated looking at the ability of screening mammography to prevent death from breast cancer. I’m not going to dig into all the data now, but I want to mention the most recent Cochrane Review (the “Holy Grail” of Evidence Based Medicine (EBM)) and the 2012 New England Journal of Medicine (NEJM) article that I mentioned in my last post.
Here is an excerpt from the 2011 Cochrane Review (emphasis mine):
…for every 2000 women invited for screening throughout 10 years, one will have her life prolonged and 10 healthy women, who would not have been diagnosed if there had not been screening, will be treated unnecessarily. Furthermore, more than 200 women will experience important psychological distress for many months because of false positive findings. It is thus not clear whether screening does more good than harm. [1]
And from the NEJM (emphasis mine):
Despite substantial increases in the number of cases of early-stage breast cancer detected, screening mammography has only marginally reduced the rate at which women present with advanced cancer. Although it is not certain which women have been affected, the imbalance suggests that there is substantial overdiagnosis, accounting for nearly a third of all newly diagnosed breast cancers, and that screening is having, at best, only a small effect on the rate of death from breast cancer. [2]
So the eminent minds in evidence based medicine think that it’s unclear if mammograms do more harm than good? That certainly isn’t the public message that most of us have heard…
Liars, damn liars, and statisticians
Part of the difficulty of understanding the benefits (and the risks) of mammography is understanding the statistics. Unfortunately, despite being taught some basics in medical school, I fear that many med students and physicians aren’t good at interpreting data. Indeed, a 2009 paper found that the vast majority of ob/gyns couldn’t accurately calculate the positive predictive value of a positive mammogram [3].
Even if a physician is statistically literate, data can appear much more or less convincing depending on how it’s presented. A 2011 article entitled “There is nothing to worry about”: Gynecologists’ counseling on mammography” gives some excellent examples [4]. Working with data published in 1996 from a randomized study conducted in Sweden, they emphasize the difference in absolute risk reduction and relative risk reductions. In the 1996 study, for every 1000 women that were screened there was a decrease in breast cancer deaths from four to three women in favor of the screened group. An absolute reduction in breast cancer deaths of 1 woman for 1000 screened does not sound particularly impressive, but the relative statistic of “a 25% decrease in mortality” sounds worthwhile [5]. [It is also worth noting that according to the Cochrane review above, the reduction in breast cancer mortality with screening mammography is actually 1 in 2000, or a 15% decrease in relative mortality or a 0.05% decrease in absolute mortality.]
When the data is presented as relative risk reduction and not absolute risk reduction, screening mammography looks a lot more beneficial. Interestingly, the risks of mammography (those of overdiagnosis and over treatment) are often presented as absolute rather than relative risks, seemingly downplaying the adverse consequences while exaggerating the benefits.
It’s not just relative…
Other mammography statistics can also be used to skew the perception of benefits. One statistic that has largely fallen out of favor, because of loud protestation from those calling for a realistic analysis of the benefits of mammography, is “survival statistics”.
To understand survival statistics we much first understand “lead time” and “lead time bias”. Wikipedia does a good job explaining this phenomenon, but for those that don’t want to take the time to click over- I will briefly expand.
Imagine a disease that kills a person at 65. Imagine that the person becomes symptomatic for that disease at 63, but with the use of a screening tool we can detect (but not cure) that disease at 55. The “diagnosis” is given when the disease is first detected, so the person diagnosed at 63 dies 2 years after diagnosis. The person whose disease was identified at 55 “survives” for 10 years, which sounds great- except really there is no difference in total life expectancy. Similarly, if you detect a “disease” that would never kill in the first place you can have stunning survival data…
Side note: The cancer that isn’t
No one questions that breast cancer kills. The problem is that “breast cancer” is not a single entity, and some of the things that are classified as breast cancer aren’t even in the same ballpark as the diseases that kill. Case in point is Ductal Carcinoma in situ (DCIS). Despite having the word “carcinoma” in its name, calling DCIS “cancer” isn’t really fair, though it can progress to cancer. Sadly we don’t know when, why, or in whom it will progress to invasive cancer. However, in the majority of women it just sits there, in situ, and is something the woman dies (or would die, if it were left alone,) with, not from [6]. Including the diagnosis of DCIS in survival statistics further skews an already questionable statistic.
Back to stats…
Promoting mammography by saying that it increases 5-year survival from 23% to 98% sounds impressive, while the actual reduction in the chance of a woman in her fifties dying from breast cancer over the next ten years only drops from 0.53% to 0.46% with mammography [7].
Perception
If you’ve made it this far, you (like me) may be becoming underwhelmed with the evidence supporting the regular use of screening mammography (and that’s without starting to consider the financial incentives that might encourage the promotion of early and often mammography…).
Unfortunately, if I poll most of my fellow classmates, they will emphatically reply that screening mammography is a good thing. It catches cancers (yes). It saves lives (marginally). It’s highly beneficial (that’s debatable).
This sentiment is not unique amongst my classmates. A recent survey shows that over 80% of responding primary care physicians believe screening mammography to be “very effective” in reducing breast cancer mortality in women aged 50-69 [8]. Another study reported that 54% of responding physicians believe that screening mammography is “very effective” at reducing cancer mortality in women aged 40-49 [9], a population where screening mammography decreases the 10 year risk of dying from breast cancer from 0.35% to 0.3% [7]. In yet another study, none of the 20 gynecologists queried mentioned risks of mammography such as over-diagnosis and over-treatment [4].
Sentiments amongst patients are similar. A 2001 study found that only 19% of women surveyed accurately assessed screening efficacy realistically, selecting that screening reduced mortality by about 25% in women over 50 (and again, this number is probably closer to 15% according to the most recent Cochrane report, and is equivalent to 1 less death per 2000 women over ten years). 50% of the women who responded estimate that screening mammography reduced breast cancer mortality by 50-75%. Not surprisingly, women who believed that screening was effective were more likely to plan to have a mammogram [10].
Women’s sentiments towards mammography are shaped by many factors. Patients, like physicians, are largely influenced by personal experiences. “Knowing someone who survived” can largely influence personal beliefs, as can the media and statements from celebrities and politicians. The type of media a woman gets her information from can also largely influence her perspective. A 2001 paper found that publications aimed towards women with lower education levels published articles that were clearly persuasive or prescriptive for screening mammography, while publications aimed towards more educated women included more balanced and informative messages [11]. Therefore, perhaps it is not surprising that higher levels of education are associated with more realistic expectations of mammography [12].
So what’s the Cliff-Notes version
Despite what many of us have come to believe, screening mammography is not womankind’s salvation in pink. Alas, it appears that survival (as in real survival, not a 5 year statistic) is basically unchanged whether women participate in screening mammography or not. Women that do participate also face the sizable risk of experiencing negative repercussions from mammography: false positives (being told there’s something there when there’s not- this is particularly prevalent in younger populations), over diagnosis, and over treatment.
I don’t want to downplay breast cancer. Breast cancer is real. Breast cancer is terrible. Breast cancer kills. But the statistics show that whether women are screened or whether a cancer is caught with diagnostics after a lump is appreciated, population survival is largely unchanged. Furthermore, women suffer ill consequences from over diagnosis and over treatment from screening mammography.
So what should we do?
Some of the screening recommendations are heading in the right direction. While the American College of Gynecologists (ACOG) and the American Cancer Society (ACS) recommend that women initiate annual screenings at the age of 40, the most recent US Preventative Task Force (USPTF) recommendations recommend starting biennial mammograms at 50.
Personally, I think the USPTF is heading in the right direction, but I, for one, would like to see a mammography recommendation similar to the recommendations for PSA testing for men given by the American Urology Association as I wrote about in my last post. We shouldn’t do it in the young (read 40-50), we shouldn’t do it in the old (and instead of “old” we really need to talk about life expectancy), and those patients in the middle need to have a serious talk with their doctor about the risks, benefits, and their personal values.
We need personalized medicine. Instead of a carte blanche recommendation about when to start mammography, we need real discussions about an individual’s risks, their values, and the potential benefits and risks of screening. Of course- that’s a lot more difficult than handing a prescription for a mammogram to every 40 year old woman who walks through the door, but I think that as doctors, we are up to the challenge.
Of course, doctors aren’t up for the challenge if they’re only given 5 minutes to talk to a patient. We need to value primary care doctors, and the doctor patient relationship, if we’re going to make strides towards personalized medicine- the question is whether the system is up to that challenge, but that’s a question for another day.
1. Gotzsche, P.C. and M. Nielsen, Screening for breast cancer with mammography. Cochrane Database Syst Rev, 2011(1).
2. Bleyer, A. and H.G. Welch, Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med, 2012. 367(21): p. 1998-2005.
3. Gigerenzer, G., Making sense of health statistics. Bull World Health Organ, 2009. 87(8): p. 567.
4. Wegwarth, O. and G. Gigerenzer, “There is nothing to worry about”: gynecologists’ counseling on mammography. Patient Educ Couns, 2011. 84(2): p. 251-6.
5. Nystrom, L., L.G. Larsson, S. Wall, L.E. Rutqvist, I. Andersson, N. Bjurstam, G. Fagerberg, J. Frisell, and L. Tabar, An overview of the Swedish randomised mammography trials: total mortality pattern and the representivity of the study cohorts. J Med Screen, 1996. 3(2): p. 85-7.
6. Welch, H.G., S. Woloshin, and L.M. Schwartz, The sea of uncertainty surrounding ductal carcinoma in situ–the price of screening mammography. J Natl Cancer Inst, 2008. 100(4): p. 228-9.
7. Woloshin, S. and L.M. Schwartz, How a charity oversells mammography. BMJ, 2012. 345: p. e5132.
8. Yasmeen, S., P.S. Romano, D.J. Tancredi, N.H. Saito, J. Rainwater, and R.L. Kravitz, Screening mammography beliefs and recommendations: a web-based survey of primary care physicians. BMC Health Serv Res, 2012. 12: p. 32.
9. Meissner, H.I., C.N. Klabunde, P.K. Han, V.B. Benard, and N. Breen, Breast cancer screening beliefs, recommendations and practices: primary care physicians in the United States. Cancer, 2011. 117(14): p. 3101-11.
10. Chamot, E. and T.V. Perneger, Misconceptions about efficacy of mammography screening: a public health dilemma. J Epidemiol Community Health, 2001. 55(11): p. 799-803.
11. Dobias, K.S., C.A. Moyer, S.E. McAchran, S.J. Katz, and S.S. Sonnad, Mammography messages in popular media: implications for patient expectations and shared clinical decision-making. Health Expect, 2001. 4(2): p. 127-35.
12. Domenighetti, G., B. D’Avanzo, M. Egger, F. Berrino, T. Perneger, P. Mosconi, and M. Zwahlen, Women’s perception of the benefits of mammography screening: population-based survey in four countries. Int J Epidemiol, 2003. 32(5): p. 816-21.
principleintopractice 