Related Experiment Videos
On the efficacy of screening for breast cancer
David A Freedman1, Diana B Petitti, James M Robins
1Department of Statistics, University of California, Berkeley, CA 94720-3860, USA.
Background:
'Mammography' (screening for breast cancer by X-ray examination) came to be widely-although not universally-accepted in the 1980s when a number of clinical trials demonstrated a substantial reduction in risk. Early detection, before the disease spread, permitted therapy that was simultaneously less invasive and more effective. Questions that remained were largely about efficacy for younger women and optimal frequency for older women. The consensus was challenged in a series of papers by two researchers at the Nordic branch of the Cochrane collaboration, Gøtzsche and Olsen, who concluded that mammography does not save lives: instead, it exposes women to unnecessary surgical procedures.
Methods:
Qualitative review.
Results:
The basis for the Gøtzsche-Olsen critique turns out to be simple. Studies that found a benefit from mammography were discounted as being of poor quality; remaining negative studies were combined by meta-analysis. The critique therefore rests on judgements of study quality, but these judgements are based on misreadings of the data and the literature.
Conclusion:
The prior consensus on mammography was correct.