Intended for healthcare professionals

Letters

Reduction in mortality from breast cancer

BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7498.1024 (Published 28 April 2005) Cite this as: BMJ 2005;330:1024

Decrease with screening was marked in younger age group

  1. Per-Henrik Zahl, senior statistician,
  2. Jan Mæhlen, professor (jan.mahlen{at}medisin.uio.no)
  1. Norwegian Institute of Public Health, PO Box 4404, N-0403 Oslo, Norway
  2. Department of Pathology, Ullevål University Hospital, N-0407 Oslo, Norway

    EDITOR—Based on a complicated statistical model, Olsen et al reported that the breast cancer mortality in Copenhagen dropped by 25% for women aged 50-79 when comparing mortality in the screening period (1991-2001) with a pre-screening period (1981-91) and claimed that this decrease was caused by screening.1

    According to Statistics Denmark, 217 breast cancer deaths occurred in Copenhagen in the age group 35-54 years in the period 1981-90 and 196 deaths in 1991-2000.2 By comparing these data with those in the article by Olsen et al we calculate that less than 5% of the breast cancer mortality in this age group is caused by tumours diagnosed after age 50 years when women were invited to screening. The number of exposure years increased from about 484 000 to about 538 000. This gives a significant 19% (age adjusted) reduction in breast cancer mortality in the age group 35-54 years. No reduction at all occurred in the age group 55-79.2

    Our estimates are based on raw and complete data. In contrast, the estimates in the paper by Olsen et al are sensitive to exclusion of data and model choice. In Copenhagen a drop in breast cancer mortality may reflect a higher fertility quotient due to non-Western immigration, fewer nulliparous women, and other non-screening related factors.

    It is therefore strange that no causal note was made by Olsen et al on the marked decrease in the breast cancer mortality in the age group 35-54 years.

    Footnotes

    • Competing interests None declared

    References

    1. 1.
    2. 2.
    View Abstract