Abstract
In the last decade, some researchers have questioned the benefits of mammographic breast cancer screening, focussing in particular on whether the mortality reduction is large enough to justify the harms. The ongoing debate, in academic journals as well as the lay press, has caused confusion rather than provide clear answers to women and their physicians as to whether the women should attend mammographic screening. Several experts have also expressed concern that women are not fully informed about the potential harms of screening, in particular about cancers that might not have been diagnosed clinically in the absence of screening.
Overdiagnosis, and resulting overtreatment, is an inevitable side effect of screening, which results partly from mammographically detected cancers that would never become symptomatic and partly from cases, where breast cancer was detected, but the woman dies of other causes before symptoms would have developed. In these cases, a woman will only experience the harmful effects of early detection and treatment, but will not gain from early diagnosis in terms of improved prognosis. There is a vigorous search for information that will allow clinicians to recognize those cancers that would not otherwise progress. However, at the moment, it is not possible to identify these individual cases and act accordingly. The number of cases overdiagnosed can only be estimated epidemiologically at the population level, based on analysis of data collected over years of screening.
In essence, all methods to estimate overdiagnosis are based on a comparison of the incidence in screened and unscreened populations. Ideally, this comparison would take places in the context of a randomized controlled trial. However, limited data for such a comparison are available for breast cancer screening since most randomized trials offered screening to the control groups at the end of the trial. In addition, randomized trial estimates refer to an experience of mammographic screening in an experimental setting over 20 years ago. Data on overdiagnosis from observational studies in current mammographic service screening programs are therefore warranted, but need to deal with various methodological challenges.
Overdiagnosis can be estimated by comparing incidence in screened and unscreened populations provided that (i) there are similar underlying risks of breast cancer in the two populations, and (ii) the effect of lead time is accounted for. Lead time refers to the period of time that a diagnosis from mammography is brought forward in time compared with clinical diagnosis. As screening is thus expected to influence incidence rate, a major difficulty in studying overdiagnosis is to disentangle the excess of incidence due to lead time from the excess due to overdiagnosis. The main approach used to deal with this issue is the cumulative-incidence approach (also called excess-incidence approach), but it should be noted that full adjustment for lead time requires very long-term follow-up. Alternatively, and especially when follow-up time in a study is limited, statistical adjustments for lead time (also referred to as the lead-time approach) have also been used.
To date, observational studies have reported a wide range of estimates ranging from 0 to 54%. Given the above, the main reason for this wide range is that the methodological approaches used differ considerably between studies, resulting from a lack of agreement on the way data should be analysed. A recent review by the Euroscreen working group showed clear differences between the estimates from studies that did or did not adequately adjust for breast cancer risk and lead time. In addition, there is a variety of measures that are used to quantify the extent of overdiagnosis, which has been shown to produce outcomes that vary by a factor of 3.5. Last, but not least, estimates are expected to differ due to differences in screening policies and different uptakes between programs.
Citation Format: Mireille J. M. Broeders. Overdiagnosis in mammographic screening. [abstract]. In: Proceedings of the Twelfth Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2013 Oct 27-30; National Harbor, MD. Philadelphia (PA): AACR; Can Prev Res 2013;6(11 Suppl): Abstract nr FO02-01.
RSS Feeds