Breast cancer screening: evidence of benefit depends on the method used

In this article, we discuss the most common epidemiological methods used for evaluating the ability of mammography screening to decrease the risk of breast cancer death in general populations (effectiveness). Case-control studies usually find substantial effectiveness. However when breast cancer mortality decreases for reasons unrelated to screening, the case-control design may attribute to screening mortality reductions due to other causes. Studies based on incidence-based mortality have obtained contrasted results compatible with modest to considerable effectiveness, probably because of differences in study design and statistical analysis. In areas where screening has been widespread for a long time, the incidence of advanced breast cancer should be decreasing, which in turn would translate into reduced mortality. However, no or modest declines in the incidence of advanced breast cancer has been observed in these areas. Breast cancer mortality should decrease more rapidly in areas with early introduction of screening than in areas with late introduction of screening. Nonetheless, no difference in breast mortality trends has been observed between areas with early or late screening start. When effectiveness is assessed using incidence-based mortality studies, or the monitoring of advanced cancer incidence, or trends in mortality, the ecological bias is an inherent limitation that is not easy to control. Minimization of this bias requires data over long periods of time, careful selection of populations being compared and availability of data on major confounding factors. If case-control studies seem apparently more adequate for evaluating screening effectiveness, this design has its own limitations and results must be viewed with caution. See related Opinion article: http://www.biomedcentral.com/1741-7015/10/106 and Commentary http://www.biomedcentral.com/1741-7015/10/164