Can we say who has died because of the breast-screening appointment error?

Newspapers are claiming to have identified some of the ‘up to 270’ women who Jeremy Hunt claimed may have died early because of the breast-screening programme error. This seems both inaccurate and hurtful.

The current NHS Breast Screening leaflet, which women receive when they are offered an appointment for screening, contains some fairly complex information based on the Marmot review of breast screening. The numbers can be summarised using the graphic below (although annoyingly such an image does not appear in the leaflet).

What might be expected to happen to 200 women who either do or do not attend breast screening every three years between ages 50 and 70, and then are followed up until aged 80. Graphic by Mike Pearson, from information in NHS Screening leaflet.

The leaflet makes clear that screening has harms as well as benefits.

Screening saves about 1 life from breast cancer for every 200 women who are screened. This adds up to about 1,300 lives saved from breast cancer each year in the UK.
About 3 in every 200 women screened every 3 years from the age of 50 to 70 are diagnosed with a cancer that would never have been found without screening and would never have become life-threatening. This adds up to about 4,000 women each year in the UK who are offered treatment they did not need.
Overall, for every 1 woman who has her life saved from breast cancer, about 3 women are diagnosed with a cancer that would never have become life-threatening.

But additional conclusions can also be read off the graphic.

The first is relevant to those who believe that screening saved their lives:

  • Out of every 12 women who survive their breast cancer after it was detected at screening, 11 would have survived even if they had not gone to screening (and 3 of whom would never had known they had breast cancer).

The second is crucial to those now trying to assign blame.

  • Out of every 4 women who missed screening and subsequently died of breast cancer, 3 would have died of their cancer anyway, even if they had gone to screening.

People have undoubtedly been harmed by this coding error (although it seems extremely implausible that 270 women have died already as a consequence). But these are ‘statistical’ harms — I don’t see how it will be possible to say with any confidence that any particular individual has died because of this error. For newspapers (or lawyers) to suggest otherwise only seems to make a bad situation even worse.

Like what you read? Give David Spiegelhalter a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.