Academics at Harvard School of Public Health looked at nearly 40,000 cases of breast cancer in Norway. They took advantage of a natural experiment to determine the effect of screening as different regions in Norway introduced screening at different times between 1996 and 2005.
The findings, presented in Annals of Internal Medicine, showed 15% and 25% of breast cancers were “overdiagnosed” by screening.
This worked out at preventing one death for every 2,500 women screened, but six to 10 cases of overdiagnosis.
Areas to think about:
- By what criteria should the effectiveness of screening programme be measured?
- What are some of the problems caused by false positives, or indeed true positives that still do not require treatment?
- What threshold should a test meet to be considered cost effective and ethical to implement?
- What kinds of biases and pressures make it hard to cut down a screening programme, once it’s already estabilished, even if evidence were to suggest it is not as effective as first thought?