Many of us will know someone who had breast cancer found at screening. The cancer seen on a mammogram may have been microscopic, so early in its malignant life that it hadn’t broken through the wall of its milk duct. Thank heavens, then, for breast screening, which is offered to all women between 50 and 70 in the UK and other countries. It is promoted enthusiastically as lifesaving, but does it deliver on its promise? And are the randomised controlled trials that persuaded governments to offer it, carried out in the 1970s and 80s, still valid?
The solution
Yes, but mostly no, seems to be the answer. Screening for cancer only works if it reduces deaths from cancer. It does this by reducing the number of the advanced cancers that are more likely to kill people. What screening doesn’t want to do is just identify breast cancers earlier (so someone has the diagnosis for longer) or find breast cancers that wouldn’t have caused any trouble in that person’s lifetime. A study of Dutch women in the BMJ concludes that rates of more advanced breast cancers didn’t fall between 1989 (when screening started) and 2012, and that while more women survived for longer, about 28% of that was due to better treatment and only 5% or less due to screening. Supporting this finding was the fact that women under 50 who weren’t eligible for screening had a similar fall in death rates over the same period.
The researchers found that more than half the cancers identified by screening would never otherwise have been found or caused problems. These women had breast surgery and anxiety that they didn’t need. The researchers pointed out that Dutch and Belgian cancer deaths both fell by 34% in the same period – although the Belgians did not start their screening programme until 15 years later.
The Welsh NHS breast screening advice says that screening reduces your risk of dying from breast cancer by 35%. But it is more meaningful to look at how many women would need to be screened to prevent deaths from breast cancer. Michael Baum, a professor emeritus of surgery at University College London, estimated that 10,000 women would need to be screened to prevent three to four deaths and that this would lead to 120-140 women being overdiagnosed and having unnecessary treatment. In an editorial accompanying the Dutch research, Mette Kalager, an associate professor from Harvard, writes: “The good news is that breast cancer-related mortality is falling. The bad news is that screening mammograms are unlikely to be responsible for that benefit while causing well-documented harm.” So far, I have not had screening, but it’s an individual choice and hard to resist.
