Seven studies (728,642 women) were included in the review.
The included studies all suffered from a number of methodological limitations, especially the possibility of selection bias. Other flaws included the misclassification of screening mammography (confused with diagnostic mammography), misclassification of death from breast cancer and lack of an adjustment for potential confounders.
Effect of screening on mortality (2 retrospective cohort studies, 14,953 women, and 1 case-control study, 198 women).
One cohort study showed a higher risk of death from breast cancer among women aged 75 to 84 years who had not been regularly screened for breast cancer compared with those who had (hazard ratio, HR 2.47, 95% confidence interval, CI: 1.70, 3.58). An increased risk was also observed in women aged over 85 years, although this increased risk was not statistically significant (HR 1.45, 95% CI: 0.63, 3.32). The second cohort study showed that women with mammographically detected tumours and no co-morbidity had lower relative risks of all-cause death compared with women whose tumours were detected clinically (age 75 to 79: relative risk, RR 0.36, 95% CI: 0.26, 0.49; age 80 to 84: RR 0.66, 95% CI: 0.52, 0.83). The findings were similar for women with moderate co-morbidity. The case-control study found no significant effect on mortality of attending screening following the last invitation, compared with not attending screening, in women aged over 75 years (RR 2.87, 95% CI: 0.62, 13.2).
Effect of screening on stage of disease at diagnosis (4 retrospective cohort studies, 713,491 women).
All studies found that women who underwent screening mammography had significantly smaller tumours and earlier disease stage at diagnosis compared with those who were not screened. Some studies found that mammography may also identify clinically insignificant tumours.