Friday, 11 May 2012

Cancer screening part 1: mammography

Cancer screening is a no brainer for most people, who think: “why wouldn’t you do it?” It turns out that there are lots of reasons why you wouldn’t do it; reasons that (of course) fall under the all too familiar heading of Overestimating the Benefits and Underestimating the Harms. Time to take a look at the risk-benefit balance.

Debate has raged over mammography for breast cancer screening, despite general acceptance from the public, and active promotion from the branches of the medical-industrial complex that depend on breast cancer screening for their business.

See here for some interesting articles against breast cancer screening from the British Medical Journal and Australian data. But there are just as many studies like this concluding that mammography is useful. In order to put it to the real test, we should randomise people to two groups: mammography screening and no routine mammography screening, and follow the groups over many years. Well, that is what was done, and the highest level of evidence we have about mammography comes from the Cochrane review of those trials. The authors reviewed seven trials involving over 600,000 women from different countries.

A side note. The usual outcome measure for these studies is “disease-specific mortality” (which I disagree with) rather than overall mortality (whether you lived or died – a much more patient-relevant outcome if you ask me). The problem with the difference is that with screening, you diagnose so many more cases, particularly ones that would never cause a problem, that the proportion of breast cancer cases that die is much lower in the screening group (raising the denominator lowers the percentage). Even if you only look at the disease-specific mortality (those that died of breast cancer, divided by the total number in that group), you are ignoring the possibility of deaths that occurred due to screening – either from the radiation itself, or the overtreatment that occurs with the inevitable “false-positive” rate you get from screening.

What did they find? Well, as you would expect, the disease specific mortality (deaths from breast cancer) was 15% less in the group that were randomised to receive screening.

What you would NOT expect is that the overall mortality (your chance of dying, from anything) was exactly the same, whether you were in the screening group or not.

What is also underreported by promoters of breast cancer screening, is that for every 2000 women screened, 200 were falsely diagnosed (false positives) which caused psychological distress, and 10 were treated for cancer unnecessarily. In the studies, thousands of women in the screening group had invasive, psychologically and physically harmful treatment for no benefit.

But don't worry, plain old mammography is old news, the credulous forward thinkers in the medical world are already onto the Next Big Thing: MRI for breast cancer screening. This will solve the problem, as it is high-tech, very expensive, and will pick up a lot more possible cancers and harmless lumps than mammography alone. And you thought the rates of false positive diagnosis and over treatment were high now?

Addit 25 July 2012: Blog post from 'The NNT', with good coverage of this debate (here).
Addit 13 Jan 2014: Screening leads to increased diagnosis due to overdiagnosis not early detection (here).
Addit 15 Feb 2014: 25 year follow up of randomised Canadian trial of mammography in 90,000 women: no difference in mortality between groups, overdiagnosis in the mammography group (here)

No comments:

Post a Comment