The benefits and harms of breast cancer screening.

The Independent UK Panel on Breast Cancer Screening (Nov 17, p 1778) estimates that one woman will avoid dying from breast cancer for every three women who are overdiagnosed. This estimate is far too positive, and the panel did not pay attention to important issues in the testimony we submitted to it in February, 2012. The panel used the data from our Cochrane review, but did not consider it important that some trials are more reliable than others, and estimated a 20% eff ect of screening. On the basis of assumptions that this eff ect would be the same today and would exist undiminished up to age 79 years, 10 years after screening stopped, the panel estimated that screening prevents about 1300 breast cancer deaths every year in the UK. These are serious errors. Contrary to the opinion of the panel, the important advances in treatment that have occurred since the trials were done will reduce the eff ect of screening substantially. A woman who would have died without screening in the past might now live so much longer, because of better treatment, that she dies of a heart attack. Screening can have no eff ect for such women. Breast cancer awareness has also reduced the eff ect of screening. The panel does not think that adjudication of the cause of death was a problem in the trials. We have documented at length, in our Cochrane review and in our report to the panel, that this is a huge problem, which inevitably biases the trials in favour of screening even when blinded endpoint committees have been used. The panel noted that all-cause mortality is not an appropriate outcome for trials of breast screening because the trials were not designed with suffi cient power for this outcome. Whether an outcome is appropriate or not has nothing to do with power. What matters is whether the outcome is reliable, and since mortality from breast cancer is not reliable, we need to look at other mortality outcomes. Screening did not reduce total mortality, nor mortality from cancer, including breast cancer (relative risk 1·02, 95% CI 0·95–1·10). It is also important to be aware that some of the healthy overdiagnosed women will die from their treatment. For example, radiotherapy increases deaths from heart disease by 27%. The panel based their 19% estimate of overdiagnosis on trials that ran for only 7–9 years, although screening is off ered for 20 years. The estimate of overdiagnosis in the Cochrane review was 29%, and observational studies have found 33% overdiagnosis in Denmark (which has an ideal control group because 80% of the country was not screened for 17 years) and 52% in a systematic review of countries with organised screening programmes. Is it acceptable that a public health initiative each year converts thousands of healthy women into cancer patients unnecessarily, which is fatal for some of them?