Mammographic screening for breast cancer. Overdiagnosis: an insidious adverse effect of screening

Mammographic mass screening for breast cancer sometimes detects tumours that would grow so slowly that the women concerned would die from other causes before developing noteworthy symptoms. This "overdiagnosis" unnecessarily exposes women to cancer treatments and their adverse effects. To assess the risk of overdiagnosis during mammographic mass screening for breast cancer, we conducted a review of the literature using the standard Prescrire methodology. The most reliable estimates of overdiagnosis are obtained by deduction, based on data from randomised trials with lengthy post-screening follow-up. A randomised trial in Sweden and two in Canada compared screening versus no screening, and monitored the women for an average of 15 and 17 years after the end of screening. For every 10 000 women who were screened, there were 63 cases of breast cancer overdiagnosis in the Swedish trial and 26 in the Canadian trials, representing about 25% of all cancers diagnosed by screening. This rate of overdiagnosis is probably on the low side, as about 20% of the women in the unscreened control groups had at least one mammogram. Studies of the general female population confirm that mammographic mass screening leads to overdiagnosis, but the results add little to those of randomised trials. In 2015, the most reliable estimate is that at least 25% of breast cancers detected by screening mammography are overdiagnosed. In France, at least 19 breast cancers would be overdiagnosed for every 1000 women screened during a 20-year period, corresponding to a total of about 3800 overdiagnosed cancers for each year of mass screening. Women must be informed of the risk of overdiagnosis and its consequences in terms of unnecessary treatment, along with other factors influencing the harm-benefit balance. See page 190 for our proposals on how to inform women wondering whether or not to undergo mammographic screening.