Evaluation of screening mammography effectiveness: The IARC recommendations of 2015 need revision

Abstract

Breast cancer screening effectiveness is measured by its ability to reduce breast cancer mortality in real world settings. A viewpoint issued by the International Agency for Research on Cancer (IARC) in 2015 considered that observational studies (cohort and case-control) are the preferred methods to evaluate the effectiveness of mammography screening. A statistical method was deemed to correct for differences in personal characteristics between women choosing to attend or to not attend screening (self-selection bias). But we showed in a systematic review of observational studies that, compared to women who never attend screening, attenders have a 45 % reduction in their risk of breast cancer death (95 % CI: 40–50 %) and a 46 % reduction (95 % CI: 42–50 %) in their risk of death from a cause other than breast cancer. Screening mammography has no known beneficial influence on causes of death other than breast cancer. If screening mammography reduced the risk of breast cancer death, one would expect a much greater percentage reduction in the risk of breast cancer death than in the risk of all-cause death. Our results suggest that screening mammography attendance is an indicator of personal characteristics associated with a lower risk of dying from any cause, including from breast cancer, which observational studies have misinterpreted as a screening effect. This paper calls into question the IARC 2015 recommendations and advocates the reinstatement of recommendations made by the IARC in 2002 which were based on the use of descriptive and comparative studies.