The debate over breast cancer screening is not for the faint-hearted. Professor Michael Baum, an eminent cancer surgeon who helped set up the UK cancer screening programme, told me once about what happened when he spoke to an audience of women in the US about his doubts about mammography. ‘They threw chairs at me,’ he said, seemingly still baffled by the angry reaction.
So please, no chairs. It’s a complicated field and I’m going to do my best to pick my way through the recent research to give some kind of answer to the question.
The debate hotted up in the last 2 weeks with the publication of two different papers by two different research teams, looking at the impact of breast cancer screening on deaths from breast cancer. The research teams came to wildly different conclusions. The first team said screening made no difference to the numbers of women dying from breast cancer. The second said thousands of lives had been saved. How did they reach such different conclusions?
To begin with, they looked at different populations. The first team looked at what happened in Denmark, where breast cancer screening was introduced in some districts in the early 1990s, but not at all in other districts. Between 1997 and 2007, deaths from breast cancer fell by 1 percent to 2 percent, in all districts. Screening didn’t make an obvious difference. (For more on this study, see my colleague Philip Wilson’s analysis on The Guardian.)
The second study, published a week later, re-analysed data from two studies, one from Sweden and one from UK. I’m going to focus on the UK study. Because breast cancer screening began simultaneously across the UK, there was no way to compare screened with unscreened populations. Instead, the researchers looked at deaths from breast cancer before and after the introduction of screening. They used a complex formula to calculate the number of fatal cancers they’d expect to see, compared to the numbers that actually happened, since the introduction of screening. They calculated that 5.7 deaths from cancer were prevented for every 1000 women screened, in the 20 years since screening began.
So it is possible that British women were more at risk of getting cancer than Danish women, and that screening saved many of these women from dying. Or its possible that the statistical team made some incorrect assumptions about how many cancer deaths we would have seen in the UK, had screening not been introduced.
Certainly and happily, deaths from breast cancer have fallen fast in the UK, although the numbers actually being diagnosed have increased 6% in the last 10 years, according to Cancer Research UK. Could this be the key? Prof Baum has said he believes the fall in the numbers of deaths is to do with improved treatment, not screening. Screening means more people are being diagnosed, but the fall in deaths is down to better surgical techniques and better drugs.
The idea of screening is to catch cancers early, while they are easy to treat and before they spread. Mammography, which uses x-rays to take a picture of the breast, can reveal lumps to small to feel or notice. The big question is, would these little lumps have grown? Most screening experts now admit that mammography leads to a certain amount of over-diagnosis, where small abnormalities are picked up by screening, although the woman would never have had any trouble from them. The question is, how much over-diagnosis?
The UK study described above says over-diagnosis is far outweighed by lives saved. The researchers claim around 2 lives are saved for every over-diagnosed case. I’ll hold my hand up to admit that, despite poring over the study, I don’t fully understand how they’ve arrived at these figures. The statistical models in the paper are complex and beyond most non-statisticians. (Any statisticians reading, please take a look and let me know what you think!)
For the sake of argument, let’s say both studies are right. Overall, the population of Denmark had no net gain in women’s lives from breast screening. And the UK has had a net gain, one that outweighs the risk of over-diagnosis. I’m using terms like ‘net gain’ for a reason. These studies still don’t answer the initial question – is it worth me, you, your mum, my auntie, going for breast screening?
I don’t know. In Denmark, there are doubtless individual women who owe their lives to having their breast cancer picked up by screening, while it was treatable. And in Britain, there are doubtless women suffering through surgery and chemotherapy, who would never have needed treatment if they’d not gone for mammography. Population-level studies are designed for public health specialists and can only tell us so much. They can’t predict what will happen to you, as an individual.
It comes down, as so often, to what you think about risk. Would you find it impossible to ignore that screening invitation, for fear of missing something that might be fatal in future? Or would you rather wait till trouble found you, blissful in ignorance? What I do think is important is that women are told the truth – or the closest we can get to it – about screening. We can’t promise that all cancers will be picked up by mammography. We don’t know if every cancer picked up this way would have caused trouble. Personally, I’m not convinced we can show for sure that cancer screening saves lives. I’m not alone – see Dr Margaret McCartney’s FT blog on the topic.
I’m still a few years away from that screening invitation dropping through my door. When it comes, I’ll think about my state of health, my personal risk factors for breast cancer, what the researchers say, and make my own decision. I don’t expect it to be easy.
Image: thanks, as always, to Wellcome Images. Credit: Annie Kavanagh.
NOTE: The Guardian’s contract with BMJ for news stories has now ended, so some of the links above no longer work. Apologies.